DANIEL SAITO

CARACTERIZAÇÃO DAS COMUNIDADES BACTERIANAS ASSOCIADAS ÀS INFECÇÕES ENDODÔNTICAS: ABORDAGEM INDEPENDENTE DE CULTIVO

Tese apresentada à Faculdade de Odontologia de Piracicaba, da Universidade Estadual de Campinas, para obtenção do Título de Doutor em Biologia Buco-Dental, Área de Concentração Microbiologia e Imunologia.

Orientador: Prof. Dr. Reginaldo Bruno Gonçalves

PIRACICABA 2007

i

FICHA CATALOGRÁFICA ELABORADA PELA BIBLIOTECA DA FACULDADE DE ODONTOLOGIA DE PIRACICABA Bibliotecário: Marilene Girello – CRB-8a. / 6159

Saito, Daniel. Sa28c Caracterização das comunidades bacterianas associadas às infecções endodônticas: abordagem independente de cultivo. / Daniel Saito. -- Piracicaba, SP : [s.n.], 2007.

Orientador: Reginaldo Bruno Gonçalves. Tese (Doutorado) – Universidade Estadual de Campinas, Faculdade de Odontologia de Piracicaba.

1. Reação em Cadeia da Polimerase. 2. RNA Ribossômico 16S. 3. Polimorfismo de Fragmento de Restrição. I. Gonçalves, Reginaldo Bruno. II. Universidade Estadual de Campinas. Faculdade de Odontologia de Piracicaba. III. Título. (mg/fop)

Título em Inglês: Characterization of bacterial communities associated with endodontic infections: culture-independent approach Palavras-chave em Inglês (Keywords): 1. Polymerase Chain Reaction. 2. 16S Ribosomal RNA. 3. Restriction Fragment Length Polymorphism. Área de Concentração: Microbiologia e Imunologia Titulação: Doutor em Biologia Buco-Dental Banca Examinadora: Marli de Fátima Fiore, Luis Eduardo Aranha Camargo, José Francisco Höfling, Sérgio Roberto Peres Line, Reginaldo Bruno Gonçalves Data da Defesa: 14-12-2007 Programa de Pós-Graduação em Biologia Buco-Dental

ii

iii

Dedico este trabalho

À minha querida esposa Cris, por todo o seu amor, a compreensão e a dedicação À nossa filhinha recém-nascida Marieva, que trouxe nova inspiração às nossas vidas Aos meus pais Wilson e Tsai, pelo amor incondicional de uma vida inteira À minha irmã Lin, por todo o carinho e a amizade compartilhados desde a nossa infância Aos meus avós, que sempre me acompanham na mente e no coração

iv AGRADECIMENTOS

Ao Prof. Dr. Reginaldo Bruno Gonçalves, grande amigo e orientador, cuja experiência e sabedoria guiaram meus passos no doutorado, contribuindo imensamente para a minha formação acadêmica.

Ao Prof. Dr. José Francisco Höfling, chefe do Laboratório de Microbiologia e Imunologia, amigo de longa data e que sempre me apoiou nos momentos mais difíceis dessa caminhada.

Aos amigos do Curso de Pós-Gradução em Biologia Buco-Dental Alessandra Castro Alves, Ana Paula Amoras, Bruna de Araújo Lima, Cristiane Duque, Fernando Zamuner, Flávia Sammartino Mariano, Gustavo Obando Pereda, Iza Alves Peixoto, Janaína de Cássia Orlandi Sardi, Marlise Inêz Klein, Paula Cristina Aníbal, Rafael Nóbrega Stipp, Regianne Umeko Kamiya, Rita de Cássia Mardegan, Ruchelle Nogueira, Sérgio Eduardo Braga Cruz, Thaís de Cássia Negrini, Vivian Furletti e aos técnicos do Laboratório de Microbiologia e Imunologia Anderson Laerte Teixeira e Wilma C. Ferraz, pela grande amizade e pelos momentos de alegria e descontração durante toda a nossa convivência.

À Prof. Dra. Siu Mui Tsai, chefe do Laboratório de Biologia Molecular e Celular do CENA – USP, por ter cedido espaço precioso em seu laboratório durante o período de dois anos para as análises de seqüenciamento de DNA.

À minha querida esposa Cristiane Pereira Borges Saito, por ter me auxiliado na coleta de amostras clínicas.

À doutoranda Fabiana de Souza Cannavan e ao Dr. José Elias Gomes e ao Dr. Jorge Luiz Mazza Rodrigues, pelo suporte dado às técnicas de PCR, clonagem e seqüenciamento de 16S rDNA.

Ao Prof. Renato de Toledo Leonardo, por ter concedido espaço na Clínica de Endodontia da Faculdade de Odontologia de Araraquara – UNESP para a coleta de amostras clínicas e, principalmente, pela amizade e confiança sempre depositadas em minha pessoa.

v Ao Prof. Dr. Alexandre Augusto Zaia, por ter gentilmente concedido espaço no Atendimento de Urgência da Faculdade de Odontologia de Piracicaba – UNICAMP para a coleta de amostras clínicas.

Ao Prof. Sérgio Roberto Peres Line, pelos inúmeros ensinamentos passados a mim durante o meu doutorado e, acima de tudo, pela grande amizade construída ao longo desses anos.

Ao Prof. Dr. Luiz Lehmann Coutinho e a todos os colegas do Laboratório de Biotecnologia Animal da ESALQ – USP, por estarem sempre dispostos a dividir a infraestrutura e o conhecimento sobre a técnica de PCR em Tempo Real.

Ao Prof. Dr. Terence Lee Marsh e a todos os colegas do Center for Microbial Ecology, Michigan State University, pelos ensinamentos sobre a técnica de T-RFLP e ecologia microbiana, e por terem cordialmente recebido a mim e a minha esposa.

À Prof. Dra. Vânia Maria Maciel Melo do Centro de Ciências da Universidade Federal do Ceará e à Prof. Dra. Vivian Pellizari do Instituto de Ciências Biológicas - USP, colegas de trabalho da Michigan State University, pelo apoio e companheirismo dados a mim e a Cris durante o nosso intercâmbio.

Ao Dr. Éderson da Conceição Jesus, colega de trabalho da Michigan State University, por ter me repassado informações importantíssimas sobre a técnica de T-RFLP.

Aos integrantes da minha banca de qualificação, Prof. Dr. José Francisco Höfling, Prof. Dr. Ricardo Della Coletta e Dr. Rogério Castilho Jacinto, pelas importantes considerações feitas ao meu trabalho de doutorado.

A Faculdade de Odontologia de Piracicaba e a todos os integrantes do corpo administrativo, sem o qual esta tese de doutorado não poderia ser realizada.

A CAPES e a FAPESP, pelo apoio financeiro durante todo o meu Curso de Pós-Graduação e por possibilitarem o meu intercâmbio no exterior.

Agradeço sinceramente a todos.

vi RESUMO

A presente tese teve como objetivo a caracterização das comunidades bacterianas associadas às infecções endodônticas pelo emprego de técnicas moleculares independentes de cultivo. Ao todo, foram analisadas amostras intra-radiculares provenientes de 34 elementos dentários associados a infecções endodônticas. A análise de bibliotecas clonais de DNA ribossomal 16S (16S rDNA) permitiu a identificação de 2 a 14 filotipos bacterianos (espécies) por elemento dentário (média= 9,6), perfazendo um total de 46 filotipos distintos. Dentre estes, 9% foram considerados previamente desconhecidos e classificados taxonomicamente como novos membros da ordem Clostridiales. Espécies reconhecidamente endodônticas dos gêneros Bacteroides, Campylobacter, Eubacterium, Peptostreptococcus, Selenomonas, Treponema e Veillonella foram detectadas, assim como representantes de gêneros menos freqüentemente descritos, como Burkholderia, Filifactor e Megasphaera. O emprego da técnica quantitativa de PCR em Tempo Real, possibilitou a detecção de P. gingivalis, T. forsythia e a coexistência de ambas em 24%, 56% e 18% dos pacientes avaliados, respectivamente. Nenhuma correlação significativa foi evidenciada entre os níveis de ambas as espécies, individualmente ou em conjunto, e a presença de sintomatologia dolorosa. O uso de T-RFLP na avaliação da estrutura das comunidades bacterianas revelou um total de 123 (endonuclease HhaI) e 122 (endonuclease MspI) fragmentos de restrição terminais (T-RFs) distintos, com médias de 20,8 e 20,0 T-RFs por elemento dentário, respectivamente. Aproximadamente 50% dos fragmentos detectados apresentaram-se, no máximo, em 2 pacientes, indicando uma alta variabilidade na composição microbiana. As análises de clusterização e de estatística multivariada não revelaram diferenças significativas nas comunidades bacterianas entre os grupos de estudo assintomático, sensível ao toque e sintomático. De modo geral, os resultados obtidos reiteraram o conceito de que a microbiota associada às infecções endodônticas é essencialmente polimicrobiana, altamente variável entre indivíduos, e constituída predominantemente por bactérias anaeróbias Gram-positivas do filo Firmicutes. As espécies P. gingivalis e T. forsythia, embora relativamente prevalentes nas infecções endodônticas, não apresentaram correlação significativa com o desenvolvimento de sintomatologia dolorosa. Por fim, a ausência de agrupamentos de perfis bacterianos quanto aos parâmetros sintomatológicos sugere que a estrutura das comunidades bacterianas intra-radiculares não possui influência significativa no desenvolvimento da dor de origem endodôntica.

Palavras-chave: infecção endodôntica, 16S rDNA, biblioteca clonal, PCR em Tempo Real, T-RFLP.

vii ABSTRACT

The objective of the present study was to characterize the bacterial communities associated with endodontic infections by use of culture-independent molecular techniques. Overall, 34 intraradicular samples from teeth harboring endodontic infections were evaluated. 16S ribosomal DNA (16S rDNA) clone library analysis allowed the identification of 2 to 14 bacterial phylotypes (species) per tooth (mean= 9.6), with a total of 46 distinct phylotypes. Among the latter, 4 (9%) were considered previously unreported and further taxonomically classified as members of the order Clostridiales. Well-known endodontic representatives of Campylobacter, Eubacterium, Peptostreptococcus, Selenomonas, Treponema e Veillonella were detected, as well as members of less frequently reported genera, such as Burkholderia, Filifactor and Megasphaera. The application of the Real Time PCR technique permitted the detection of P. gingivalis, T. forsythia and a coexistence of both in 24%, 56% e 18% of the subjects, respectively. No significant correlations were evidenced among the levels of P. gingivalis and T. forsythia, individually or conjointly, and spontaneous endodontic pain. The use of T-RFLP in the analysis of bacterial community structures revealed a total of 123 (HhaI endonuclease) and 122 (MspI endonuclease) distinct terminal restriction fragments (T-RFs), with 20.8 and 20.0 mean T-RFs per tooth, respectively. Approximately 50% of the detected fragments were exclusive to one or two patients, indicating a high inter-subject variability in the bacterial assemblages. Cluster and multivariate statistical analyses did not demonstrate significant differences in the bacterial community profiles among the asymptomatic, tender to percussion and symptomatic study groups. Taken together, the results of this study reiterate the concept that the microbiota associated with endodontic infections is essentially polymicrobial, highly variable among individuals, and predominantly composed of Gram-positive anaerobic bacteria from the phylum Firmicutes. The species P. gingivalis and T. forsythia, although relatively prevalent in root canal infections, did not present significant correlations with the development of symptomatic features. Lastly, the absence of clusters of bacterial profiles according to symptomatic parameters suggests that the intraradicular bacterial community structures, as a whole, do not bear significant influence on the development of pain of endodontic origin.

Keywords: bacteria, endodontic infection, 16S rDNA, clone library, Real Time PCR, T-RFLP.

viii SUMÁRIO

1. INTRODUÇÃO 1

2. PROPOSIÇÃO 4

3. CAPÍTULO 1 5

4. CAPÍTULO 2 19

5. CAPÍTULO 3 35

6. CONSIDERAÇÕES GERAIS 51

7. CONCLUSÃO 53

8. REFERÊNCIAS 54

9. ANEXOS 58

ix INTRODUÇÃO

O objetivo principal da terapia endodôntica é a eliminação dos microrganismos patogênicos e de seus subprodutos do interior dos canais radiculares, bem como a manutenção da condição asséptica pós-operatória (Bystrom & Sundqvist, 1981). Mesmo com os recentes avanços técnico-científicos da Odontologia moderna, a permanência de células bacterianas no interior dos canais é, ainda, um dos principais motivos de fracasso no tratamento endodôntico (Lin et al., 1992). Nesse sentido, o conhecimento da microbiota associada às infecções endodônticas constitui requisito fundamental para o desenvolvimento de medidas terapêuticas mais efetivas.

Anteriormente à década de 1970, as dificuldades existentes com relação à cultura e à identificação de organismos anaeróbios mascarava a sua real importância na etiopatogenia das alterações pulpares e periapicais de origem infecciosa (Brown & Rudolf, 1957; Grossman, 1959; Melville & Birch, 1967). Com a subseqüente evolução das técnicas de cultivo microbiano em ambiente anaeróbio, o enfoque microbiológico na Endodontia sofreu grande mudança. A microbiota endodôntica, que antes era considerada basicamente composta por microrganismos aeróbios facultativos, demonstrou abrigar uma grande proporção de anaeróbios estritos (Gomes et al., 2004; Sundqvist, 1992; Fabricius et al., 1982; Sundqvist, 1976; Kantz & Henry, 1974). Atualmente, consideram-se as infecções endodônticas essencialmente polimicrobianas, e constituída predominantmente por bactérias anaeróbias dos gêneros Actinomyces, Campylobacter, Eubacterium, Fusobacterium, Peptococcus, Peptostreptococcus, Porphyromonas, Prevotella, Streptococcus, Veillonella, dentre outras.

Embora a microbiota associada às infecções endodônticas encontre-se amplamente caracterizada por estudos de cultivo microbiano, estima-se hoje que aproximadamente 50% das bactérias orais não seja capaz de crescer sob condições laboratoriais (Paster et al., 2001; Aas et al., 2003). Uma vez que a microbiota endodôntica representa uma parcela restrita da microbiota presente na cavidade oral (Sunqvist, 1994), é razoável considerar que parte significativa da primeira encontra-se potencialmente subestimada pelos métodos de detecção tradicionais. Sob este novo paradigma, abordagens moleculares independentes de cultivo oferecem uma alternativa rápida e sensível, permitindo também a caracterização das frações fastidiosa e não- cultivável da microbiota endodôntica (Munson et al., 2002; Rolph et al., 2001).

A análise de bibliotecas clonais consiste na amplificação do gene ribossomal 16S (16S rDNA) pela Reação em Cadeia da Polimerase (PCR), ligação dos fragmentos amplificados em vetor plasmidial, clonagem em Escherichia coli, e seqüenciamento do DNA. As seqüências nucleotídicas são submetidas ao alinhamento múltiplo e comparadas juntamente com seqüências

1 de bancos de dados de DNA ribossomal, possibilitando a identificação filogenética das espécies bacterianas (filotipos) presentes na amostra. A técnica demonstra grande sensibilidade, gerando resultados bem mais informativos que a cultura microbiana (Kroes et al., 1999; Rolph et al., 1991), com excelente aplicabilidade nos mais diversos microambientes orais, como língua, palato, placa sub e supra-gengival, abscessos dento-alveolares, lesões de cancro e canais radiculares (Aas et al., 2003; Paster et al., 2002; Paster et al., 2001; Munson et al., 2002; Rolph et al., 2001; Wade et al., 1997).

As bactérias Porphyromonas gingivalis e Tannerella forsythia possuem fatores de virulência amplamente caracterizados (Inagaki et al., 2006; Holt et al., 1999), sendo reconhecidamente importantes na etiologia da doença periodontal (Socransky et al., 2002; Socransky et al., 1998). Estudos em modelos animais demonstraram que P. gingivalis e T. forsythia possuem atividades patogênicas sinergísticas, quando inoculadas em conjunto (Yoneda et al., 2001; Takemoto et al., 1997). Embora ambas já tenham sido relatadas em infecções endodônticas através da detecção estritamente qualitativa (Foschi et al., 2005; Rôças et al., 2001; Fouad et al., 2002), a aplicação de técnicas quantitativas pode trazer informações adicionais sobre a importância patogênica destas espécies nesse ecossistema.

A PCR em Tempo Real (Real Time PCR) é uma variante da PCR convencional que oferece quantificação sensível e confiável de ácidos nucléicos (Heid et al., 1996). A técnica baseia-se no monitoramento da intensidade de fluorescência oriunda da reação de PCR, a qual é diretamente proporcional aos níveis do DNA amplificado. Uma vez que os produtos da reação são detectados e quantificados diretamente no equipamento, a técnica dispensa o processamento posterior, minimizando potenciais erros de análise (Bustin, 2000). Estudos orais demonstraram que PCR em Tempo Real possui sensibilidade de detecção 41 vezes superior àquela obtida pela contagem de colônias anaeróbias (Martin et al., 2002) e que, nos casos específicos de P. gingivalis e T. forsythia, fornece aumentos de 36 a 51% nas respectivas taxas de detecção (Verner et al., 2006).

A análise de Polimorfismos de Comprimento de Fragmentos Terminais de Restrição (T- RFLP) é uma técnica baseada na amplificação de DNA, a qual tem sido empregada para se investigar a estrutura de comunidades microbianas complexas (Marsh, 2005; Blackwood et al., 2003; Osborn et al., 2000). Consiste na amplificação de 16S rDNA polimicrobiano por PCR, onde um dos primers é marcado por fluorescência na sua porção 5’. A subseqüente digestão com endonucleases tetraméricas produz fragmentos terminais de restrição, que são precisamente mensurados em seqüenciadores automatizados (Liu et al., 1997). Uma vez que diferentes populações microbianas possuem sítios de restrição distintos, um perfil (fingerprint) genético é obtido para cada comunidade analisada. A técnica já foi empregada na comparação entre a 2 microbiota de pacientes saudáveis e periodontais (Sakamoto et al., 2003), na determinação de mudanças na estrutura de comunidades bacterianas sub-gengivais pós-tratamento (Sakamoto et al., 2004), no estudo da influência da qualidade de restaurações na composição da microbiota endodôntica (Hommez et al., 2004) e na análise comparativa entre as microbiotas intra-radicular e de abscessos dento-alveolares, em casos sintomáticos e assintomáticos (Sakamoto et al., 2006).

As técnicas moleculares supramencionadas abrigam, como característica comum, a possibilidade de detecção de bactérias sem a necessidade do crescimento laboratorial de microrganismos provenientes das amostras. Isso evita que os pesquisadores entrem em contato direto com patógenos humanos em potencial, garantindo maiores níveis de biossegurança. Ademais, uma vez que muitas das bactérias orais são reconhecidamente de crescimento lento, o uso das técnicas moleculares reduz significativamente o tempo experimental (Siqueira et al., 2002). Dentre as vantagens adicionais oferecidas pelas técnicas de detecção moleculares, destacam-se o processamento de amostras em larga escala e as altas reprodutibilidade, confiabilidade e sensibilidade (Marsh et al., 2005; Heid et al., 1996; Rolph et al., 2001; Bustin, 2000; Morrison et al., 1988).

3 PROPOSIÇÃO

A presente tese teve como objetivo o estudo da diversidade bacteriana em infecções endodônticas, por meio de técnicas moleculares independentes de cultivo laboratorial. Para tal, foram avaliados 34 elementos dentários portadores de infecções endodônticas, provenientes de diferentes pacientes. O estudo foi dividido em três fases, correspondentes às diferentes metodologias empregadas:

I. Caracterização da microbiota endodôntica, determinação da diversidade bacteriana e avaliação filogenética de filotipos (espécies) potencialmente novos, pela análise de bibliotecas clonais do gene ribossomal 16S (16S rDNA) em infecções endodônticas assintomáticas não-expostas à cavidade oral (n= 7),

II. Detecção quantitativa de P. gingivalis e T. forsythia, através da técnica de PCR em Tempo Real em infecções endodônticas assintomáticas (n= 10) e sintomáticas (n= 24), e determinação de possíveis correlações entre os níveis celulares detectados e a sintomatologia dolorosa de origem endodôntica,

III. Análise da estrutura de comunidades bacterianas associadas a infecções endodônticas, por meio da técnica de T-RFLP em infecções endodônticas assintomáticas (n= 7), sensíveis à percussão (n= 5) e sintomáticas (n= 13), e investigação de possíveis correlações entre a composição bacteriana, sensibilidade à percussão e dor espontânea de origem endodôntica.

Esta tese foi elaborada no formato alternativo, conforme deliberado pela Comissão Central de Pós-Graduação (CCPG) da Universidade Estadual de Campinas (UNICAMP) (Anexo1). Assim, as metodologias acima mencionadas (fases) encontram-se apresentadas na forma de Capítulos. Os comprovantes de publicação, submissão a periódico, apresentação de trabalho em congresso sujeito a comissão julgadora e o certificado do Comitê de Ética em Pesquisa encontram-se nos Anexos.

4 CAPÍTULO 1

Artigo publicado no periódico Journal of Medical Microbiology. 2006. 55(1): 101-107 (Anexo 2).

Identification of bacteria in endodontic infections by sequence analysis of 16S rDNA clone libraries

Authors: Daniel Saito1, Renato de Toledo Leonardo2, Jorge Luiz Mazza Rodrigues3, Siu Mui Tsai3, José Francisco Höfling1, Reginaldo Bruno Gonçalves1,4

1 Department of Oral Diagnostics, Dental School of Piracicaba, State University of Campinas, Piracicaba, São Paulo, Brazil; 2 Department of Restorative Dentistry, Dental School of Araraquara, State University of São Paulo, Araraquara, São Paulo, Brazil; 3 Laboratory of Cell and Molecular Biology, Center of Nuclear Energy in Agriculture, University of São Paulo, Piracicaba, São Paulo, Brazil

Running title: Root canal bacteria by 16S rDNA analysis Keywords: 16S rDNA, bacteria, cloning, endodontic, infection, root canal

4 Correspondence: Professor Reginaldo Bruno Gonçalves. Departamento de Diagnóstico Oral, Faculdade de Odontologia de Piracicaba, Universidade Estadual de Campinas (UNICAMP). Av. Limeira – 901, Piracicaba, 13414-903, São Paulo, Brazil. Phone: +55 19 3412-5321, Fax: +55 19 3412-5218 E-mail: [email protected]

Abbreviations: CTAB (hexadecyltrimethylammonium bromide) Footnote: The GenBank accession numbers for the 16S rDNA sequences of Uncultured Clostridiales bacterium clone AG_D03, Uncultured Clostridiaceae bacterium clone AG_G04, Uncultured Streptococcaceae bacterium clone AF_F05 and Uncultured Mogibacterium sp. clone AF_H06 are AY821867, AY821868, AY821869, AY821870, respectively.

5 Summary

A significant proportion of oral bacteria are unable to undergo cultivation by existing techniques. In this regard, the microbiota from root canals still requires complementary characterization. The present study aimed at the identification of bacteria by the sequence analysis of 16S rDNA clone libraries from seven endodontically infected teeth. Samples were collected from the root canals, subjected to the polymerase chain reaction (PCR) with 16S rDNA universal primers, cloned and partially sequenced. Clones were clustered into groups of closely related sequences (phylotypes) and identification to the species-level was performed by comparative analysis with the GenBank, EMBL and DDBJ databases, according to a 98% minimum identity. All samples were positive for bacteria and the number of phylotypes detected per subject varied from 2 (patient AB) to 14 (patient AA). The majority of taxa (65.2%) belonged to the phylum Firmicutes of Gram-positive bacteria, followed by Proteobacteria (10.9%), Spirochaetes (4.3%), Bacteroidetes (6.5%), Actinobacteria (2.2%) and Deferribacteres (2.2%). A total of 46 distinct taxonomic units was identified. Four clones with low similarities with those previously deposited in the databases were sequenced to nearly full extent and were taxonomically classified as novel representatives of the order Clostridiales, including a new species of Mogibacterium. The identification of novel phylotypes associated with endodontic infections suggests that the endodontum may still harbor a relevant proportion of uncharacterized taxa.

Introduction

Although more than 150 species of bacteria have been identified in infected root canals, only a restricted number can be found simultaneously in the same tooth and a considerable variation of species is expected when analyzing distinct clinical conditions, individuals, or populations (Sundqvist, 1976; Molander et al., 1998; Baumgartner et al., 2004). Cultivation studies have shown a predominance of facultative and strict anaerobes in the endodontum, including representatives of Eubacterium, Fusobacterium, Peptococcus, Peptostreptococcus, Porphyromonas, Prevotella, and Streptococcus (Sundqvist, 1992b; Le Goff et al., 1997). Bacteria inside the canal are the major causes of periapical pathologies (Kakehashi et al., 1965) and, if not adequately treated, can give rise to dentoalveolar abscess, a condition that has ability to initiate morbidity, life threatening illness (Walsh, 1997) and predispose to transient bacteremia during therapy (Savarrio et al., 2005). Previous reports suggested that endodontic bacteria might be involved in extra-oral complications, such as chronic maxillary sinusitis (Melen et al., 1986), orbital cellulitis (Ngeow, 1999), infective endocarditis (Bate et al., 2000), rheumatoid arthritis (Breebaart

6 et al., 2002) and brain abscess (Henig et al., 1978). In this regard, substantial understanding of the endodontic microbiota is an important requirement for both oral and medical microbiologists.

While it is common-knowledge that the development of efficient treatment strategies relies on the characterization of the endodontic microbial communities in entirety, cultivation-based techniques may cut down the range of detection, since a subset of oral inhabitants still cannot undergo cultivation (Paster et al., 2001). In this new context, clone library analysis of ribosomal DNA, particularly the 16S rDNA, has become a trustworthy tool for determining bacterial diversity, often yielding more informative results when compared to cultivation alone (Kroes et al., 1999; Rolph et al., 2001). This broad-based cultivation-free approach has been employed in the investigation of polymicrobial human infections, such as periodontal disease, childhood caries, dentoalveolar abscesses, maxillary sinusitis and noma lesions (Wade et al., 1997; Paster et al., 2001; Paster et al., 2002; Becker et al., 2002; Hutter et al., 2003; Paju et al., 2003). Particularly in root canal infections, 16S rDNA sequence analysis has enabled detection of bacteria when culture had generated negative results and has permitted the identification of novel species in relatively small sets of samples (Rolph et al., 2001; Munson et al., 2002).

Here, we report the results of an investigation on the bacterial diversity of seven infected root canals by the analysis of 16S rDNA libraries, in an effort to contribute to the ongoing characterization of the root canal microbiota.

Methods

Subjects. Seven patients, 2 males and 5 females, ranging from 15 to 42 years old (mean 27.7 ± 8.4) were analyzed. Subjects had been referred for endodontic treatment at the Dental School of Piracicaba, and were selected for presence of pulpal necrosis and chronic periapical lesions by clinical and radiographic evaluation. In order to facilitate antisepsis of the operation field during sampling, only teeth with unexposed pulp chambers were included. These were represented by teeth with intact crowns or small, clinically acceptable restorations. Subjects with periodontal pockets >3mm, advanced bone-loss, acute abscesses, tooth fractures, sinus tracts or those who had undergone antibiotic therapy within two months prior to collection were not included. Written informed consent was obtained from all individuals and ethical approval was granted by the Ethical Committee for Human Subjects of the Dental School of Piracicaba, State University of Campinas.

Sample collection. Each patient was submitted to local anesthesia and the tooth was isolated with rubber-dam. Cleaning of the tooth crown was performed to eliminate food debris and dental plaque. Antisepsis of the crown and operation field was conducted with 1.0% NaOCl for 1 min, 7 followed by inactivation with 5% sodium thiosulfate. Coronal access cavity was gained by high- speed bur irrigated with sterile saline solution. As the pulp chamber was reached, a sterile #15 K- file was introduced at 3mm short of the root apex. After careful instrumentation, the active portion of the K-file was cut and placed in a test tube containing 1 mL of TE buffer (10 mM TrisHCl, 1 mM EDTA, 0.1 mM NaCl, pH 8.0). Three sterile #15 paper-points were consecutively introduced inside the canal for 20 sec each and placed in the same test tube. Samples were immediately transported to the laboratory and stored at – 20 ºC.

DNA extraction. Samples were thawed in water bath at 37 ºC for 10 min, vortexed for 30 sec and the paper points and K-files were removed from the tubes. Bacterial cells were pelleted by centrifugation at 20,000 g for 10 min and the supernatant was discarded. A DNA extraction protocol based on chloroform: isoamyl alcohol and hexadecyltrimethylammonium bromide (CTAB) was employed (Kuipers et al., 1999; Smith et al., 1989). DNA was re-suspended in TE buffer with 10 µg RNAse ml-1, incubated in water bath at 37oC for 30 min and stored at – 20 ºC until required.

16S rDNA amplification. PCR control tests for the 16S rDNA universal eubacterial primers fD1 (5’-AGAGTTTGATCCTGGCTCAG-3’) and rD1 (5’-AAGGAGGTGATCCAGCC-3’) (Weisburg et al., 1991) were performed with 21 bacteria strains, yielding positive amplification for all DNA tested, as determined by visualization on agarose gel electrophoresis (available as supplementary data in JMM Online). Polymicrobial 16S rDNA from the clinical samples was amplified by PCR in 25 µL mixtures, as previously described (Rodrigues et al., 2003), except for the following modifications: 2 mM MgCl2 and 1.5 U Taq DNA polymerase. PCR products were examined on 1.0% low melting point agarose-gel electrophoresis stained with ethidium bromide. Amplification products (about 1,500 bp) were purified with the GFX DNA Purification kit, according to the manufacturer’s instructions (Amersham Biosciences).

Cloning of polymicrobial PCR products. 65.0 ng of each 16S rDNA PCR product was ligated to pMOSBlue vector and transformed in Escherichia coli DH10B cells, according to the manufacturer’s instructions (Amersham Biosciences). Small-scale plasmid DNA preparations were conducted by an alkaline lysis protocol as described by Sambrook et al. (1989). Screening of recombinants was performed on 1.0% agarose-gel electrophoresis stained with ethidium bromide.

16S rDNA partial sequencing. Seventy recombinant clones per library were randomly selected for partial 16S rDNA sequencing reactions, performed in a DNA thermal cycler. Temperature profile for primers T7 (5’-TAATACGACTCACTATAGGG-3’) and U19mer (5’- GTTTTCCCAGTCACGACG-3’) included an initial step at 96 ºC for 2 min, followed by 30 cycles of 96 ºC for 30 sec, 50 ºC for 30 sec and 60 ºC for 4 min. Reactions were performed in 10 µL

8 mixtures containing 250 ng of template DNA, 1 µL of Big Dye TM Terminator Ready v.3.0 (Applied Biosystems), 0.5 µM of primer and 3 µL of sequencing buffer (200 mM TrisHCl pH 9.0, 5 mM

MgCl2).

16S rDNA sequence analysis. Sequences were automatically analyzed in ABI Prism 3100 Genetic Analyzer (Applied Biosystems – Hitachi) and grouped into clusters (phylotypes), according to a 99% minimum similarity (Kroes et al., 1999; Hutter et al., 2003). One representative of each phylotype was selected and submitted to the blastn algorithm (BLAST 2.0 - http://www.ncbi.nlm.nih.gov/BLAST), allowing comparison with sequences present in the GenBank, DDBL and EMBL databases. Only the highest-scored BLAST result was considered for phylotype identification, with a 98% minimum similarity (Stackebrandt & Goebel, 1994).

Phylogenetic analysis of novel phylotypes. Four clones with BLAST identities ≤97% were considered as representatives of novel phylotypes and were sequenced with additional primers 341-357f (5’-CCTACGGGAGGCAGCAG-3’), 357-341r (5’-CTGCTGCCTCCCGTAGG-3’), 685- 704f (5’-GTAGSGGTGAAATSCGTAGA-3’), 704-685r (5’-TCTACGSATTTCACCSCTAC-3’), 1099-1114f (5’-GCAACGAGCGCAACCC-3’) and 1114-1099r (5’-GGGTTGCGCTCGTTGC-3’) (Lane, 1991). Contiguous sequences were assembled with the Phred/Phrap/Consed software package (www.phrap.org), followed by analysis with Chimera Check (RDP II - http://rdp.cme.msu.edu) to ensure inexistence of chimeric molecules. Novel phylotypes were taxonomically assigned with Naive Bayesian rRNA Classifier v.1.0 (RDP II) and submitted to phylogenetic analysis, along with close-related sequences from the order Clostridiales obtained by the Hierarchy Browser program (RDP II). All sequences were aligned with the ClustalW software (Thompson et al., 1994) and visualized with Bioedit 7.0.4 (www.mbio.ncsu.edu/BioEdit). A phylogenetic tree was constructed with MEGA 2.1 (Kumar et al., 2001), according to the calculation of a distance matrix (Jukes & Cantor, 1969) and tree reconstruction by the neighbor- joining method (Saitou & Nei, 1987). Bootstrap confidence values for branching nodes were inferred by the generation of 100 resampling trees.

Results and Discussion

The nucleotide sequence analysis of 16S rDNA clone libraries was used to investigate the bacterial diversity of seven endodontically infected teeth. All teeth evaluated were positive for the presence of bacteria. Overall, 46 taxonomic units (phylotypes) were detected (Table 1). In spite of the relative homogeneity of our study group, composed of asymptomatic teeth associated with unexposed necrotic pulps, chronic periapical lesions and no periodontal disease, a high variation in bacterial compositions could be observed: 33 (71.7%) phylotypes were subject-exclusive,

9 whereas only 13 (28.3%) could be detected in more than one patient. The number of phylotypes also ranged substantially among subjects: from 2 (patient AB) to 14 (patient AA) (mean 9.57 ±3.91). This variation is in accordance to both cultivation and molecular-based studies (Sundqvist, 1976; Jung et al., 2000). Table 1 illustrates the percentage distribution of phylotypes within each subject and is intended to provide a quantitative view on the results. This type of analysis brings important contribution to our findings (Dewhirst et al., 2000; Becker et al., 2002; Munson et al., 2002; Hutter et al., 2003), but should be viewed with discretion, since multi-template PCR can be subjected to bias in template-to-product ratios (Suzuki & Giavannoni, 1996; Polz & Canavaugh, 1998).

In general, the results of this investigation reiterated data from cultivation and molecular- based investigations, with a predominance of anaerobic bacteria, especially from the phylum Firmicutes of Gram-positives (Sundqvist, 1992a; Munson et al., 2002). Representatives of other phyla were found in much lower frequencies: Proteobacteria (10.9%), Spirochaetes (4.3%), Bacteroidetes (6.5%), Actinobacteria (2.2%) and Deferribacteres (2.2%); 9% of sequences could not be assigned at the phylum level. Common endodontic species, such as Campylobacter gracilis, Eubacterium tardum, Peptostreptococcus anaerobius, Peptostreptococcus micros and Lachnospiraceae sp. were positively identified with high percent identities (≥99%). Recently reported phylotypes were also detected: Bacteroidales oral clone MCE7_164 E2b, Lachnospiraceae oral clone MCE7_60 E1, Lachnospiraceae oral clone MCE9_173 E4, Megasphaera sp. oral clone MCE3_141 P1 (Munson et al., 2002). Dialister invisus, a newly described oral Gram-negative coccobacillus (Downes et al., 2003), was the most commonly found taxon (5 out of 7 subjects), followed by Filifactor alocis and Eubacteriaceae oral clone P2PB_46 P3 (4 out of 7 subjects). Accordingly, bacteria from the genus Dialister have been identified in oral infections with increasing frequencies (Contreras et al., 2000; Munson et al., 2002). D. pneumosintes, a species frequently associated with purulent infections, brain abscesses and bite- wounds (Goldstein et al., 1984; Rousée et al., 2002), has also been considered as putative pathogen in periodontal and endodontic infections (Ghayoumi et al., 2002; Siqueira & Rôças, 2002) and could be detected in one subject.

Phylotypes corresponding to recently proposed pathogens of periodontal disease were identified, corroborating molecular data from Paster et al. (2001) and Hutter et al. (2003). Among those, Treponema socranskii has already been detected in the endodontum, being one of the most common root canal treponemes (Baumgartner et al., 2003), while Uncultured Eubacterium clone PUS9.170 has also been found in dentoalveolar abscesses (Wade et al., 1997). Pseudoramibacter alactolyticus and Filifactor alocis were positively detected with high percent

10 matches (99%) and have proven to be frequent inhabitants of root-filled, refractory cases (Siqueira & Rôças, 2004).

Forty-two (91.3%) clones were identified to the species-level and four (8.7%) corresponded to sequences with no resemblance to any other previously deposited in the databases (Figure 1), according to the established 98% nucleotide identity threshold. This parameter is in accordance to a previously proposed species definition criterion, based on DNA-DNA re-association assays (Stackebrandt & Goebel, 1994) and lays within the range of values employed by similar studies, which vary from 98% (Sakamoto et al., 2000; Rolph et al., 2001; Munson et al., 2002) to 99% (Kroes et al., 1999; Drancourt et al., 2000; Hutter et al., 2003). Applying a 97% threshold value did not bring modifications to our findings, whereas a 99% value resulted in the detection of 15 potentially novel species, as opposed to only 4. The application of such a stringent condition was nonetheless rejected, since it could generate redundancies in the results, due to the intragenomic heterogeneities in 16S ribosomal RNA operons (Coenye & Vandamme, 2003; Acinas et al., 2004).

Among the novel phylotypes detected (Figure 1), Uncultured Mogibacterium sp. clone AF_H06 (AY821870) was the only taxonomically assigned to the genus level. Mogibacterium is a newly proposed genus of bacteria originally isolated from periodontal pockets and infected root canals, represented by anaerobic Gram-positive bacilli (Nakazawa et al., 2000). Representatives of this genus were shown to be frequent in endodontic infections, as observed by Rolph et al. (2001), who identified clones closely related to M. neglectum, M. vescum and M. diversum, with 97% nucleotide identities. The epidemiological importance of this group of bacteria in endodontic infections is still to be investigated.

Interestingly, some important endodontic species, such as and Fusobacterium nucleatum and the black-pigmented anaerobes Porphyromonas gingivalis, Porphyromonas endodontalis and Prevotella intermedia could not be detected. Lack of primer specificity was readily discounted, since we had successfully tested our universal primers with F. nucleatum ATCC 10953 and many species within the phylum Bacteroidetes, including P. gingivalis ATCC 33277, P. endodontalis ATCC and P. intermedia ATCC 25611 (data in JMM Online). Similar root canals studies were also not able to find any P. gingivalis phylotypes (Rolph et al., 2001; Munson et al., 2002). Fusobacteria have been frequently encountered in root canals by cultivation (Sundqvist, 1992a) and PCR assays (Fouad et al., 2002), but its prevalence seems to be relatively low in 16S rDNA- based studies. Accordingly, Munson et al. (2002) identified one Fusobacterium clone out of 624 sequenced, whereas Rolph et al. (2001) detected F. nucleatum solely in the refractory subset of cases. The absence of some bacteria species in the present study can be attributed to intrinsic technique limitations, or to the small sample-set investigated. 11 In conclusion, the results from this study are in accordance to those from similar research, revealing a predominance of anaerobic species from the phylum Firmicutes of Gram-positive bacteria in infected root canals, particularly from the class Clostridia. The identification of uncultured clones, originally encountered in the endodontum, saliva and subgingival plaque demonstrates, over again, that the endodontic and periodontal microbial communities may share a relevant proportion of bacteria, despite their established anatomical interrelationships (Kerekes & Olsen, 1990; Rupf et al., 2000). The identification of novel phylotypes adds to the concept that the endodontum may still harbor a relevant proportion of uncharacterized taxa.

Acknowledgements

This work was supported by grants from the Brazilian funding agencies Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP – Proc. 04/01674-6; 04/13548-5; 00/10168-6), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). We are greatly thankful to Ms. Fabiana de Souza Cannavan and Mr. José Elias Gomes for technical supports.

References

Acinas, S. G., Marcelino, L. A., Klepac-Ceraj, V. & Polz, M. F. (2004). Bias caused by template annealing in the amplification of mixtures of 16S rRNA genes by PCR. J Bacteriol 186, 2629– 2635. Bate, A. L., Ma, J. K. & Pitt Ford, T. R. (2000). Detection of bacterial virulence genes associated with infective endocarditis in infected root canals. Int Endod J 33, 194-203. Baumgartner, J. C., Khemaleelakul, S. U. & Xia, T. (2003). Identification of spirochetes (treponemes) in endodontic infections. J Endod 29, 794-797. Baumgartner, J. C., Siqueira Jr., J. F., Xia, T. & Rôças, I. N. (2004). Geographical differences in bacteria detected in endodontic infections using polymerase chain reaction. J Endod 30, 141-144. Becker, M. R., Paster, B. J., Leys, E. J., Moeschberger, M. L., Kenyon, S. G., Galvin, J. L., Boches, S. K., Dewhirst, F. E. & Griffen, A. L. (2002). Molecular analysis of bacterial species associated with childhood caries. J Clin Microbiol 40, 1001-1009. Breebaart, A. C., Bijlsma, J. W. & van Eden, W. (2002). 16-year remission of rheumatoid arthritis after unusually vigorous treatment of closed dental foci. Clin Exp Rheumatol 20, 555-557. Coenye, T. & Vandamme, P. (2003). Intragenomic heterogeneity between multiple 16S ribosomal RNA operons in sequenced bacterial genomes. FEMS Microbiol Lett 228, 45-49.

12 Contreras, A., Doan, N., Chen, C., Rusitanonta, T., Flynn, M. J. & Slots, J. (2000). Importance of Dialister pneumosintes in human periodontitis. Oral Microbiol Immunol 15, 269-272. Dewhirst, F. E., Tamer, M. A., Ericson, R. E., Lau, C. N., Levanos, V. A., Boches, S. K., Galvin, J. L. & Paster, B. J. (2000). The diversity of periodontal spirochetes by 16S rRNA analysis. Oral Microbiol Immunol 15, 196–202. Downes, J., Munson, M. & Wade, W. G. (2003). Dialister invisus sp. nov., isolated from the human oral cavity. Int J Syst Evol Microbiol 53, 1937-1940. Drancourt, M., Bollet, C., Carlioz, A., Martelin, R., Gayral, J. & Raoult, D. (2000). 16S ribosomal DNA sequence analysis of a large collection of environmental and clinical unidentifiable bacterial isolates. J Clin Microbiol 38, 3623-3630. Fouad, A. F., Barry, J., Caimano, M., Clawson, M., Zhu, Q., Carver, R. , Hazlett, K. & Radolf, J. D. (2002). PCR-Based Identification of Bacteria Associated with Endodontic Infections. J Clin Microbiol 40, 3223–3231 Ghayoumi, N., Chen, C. & Slots, J. (2002). Dialister pneumosintes, a new putative periodontal pathogen. J Periodontal Res 37, 75-78. Goldstein, E. J., Citron, D. M. & Finegold, S. M. (1984). Role of anaerobic bacteria in bite-wound infections. Rev Infect Dis 6, 177-183. Henig, E. F., Derschowitz, T., Shalit, M., Toledo, E., Tikva, P. & Aviv, T. (1978). Brain abcess following dental infection. Oral Surg Oral Med Oral Pathol 45, 955-958. Hutter, G., Schlagenhauf, U., Valenza, G., Horn, M., Burgemeister, S., Claus, H. & Vogel, U. (2003). Molecular analysis of bacteria in periodontitis: evaluation of clone libraries, novel phylotypes and putative pathogens. Microbiology 149, 67-75. Jukes, T. H. & Cantor, C. R. (1969). Evolution of protein molecules. In Mammalian Protein Metabolism, pp. 21-132. Edited by , H. N. Munro. New York, USA: Academic Press. Jung, I. Y., Choi, B. K., Kum, K. Y., Roh, B. D., Lee, S. J., Lee, C. Y. & Park, D. S. (2000). Molecular epidemiology and association of putative pathogens in root canal infection. J Endod 26, 599-604. Kakehashi, S., Stanley, H. R. & Fitzgerald, R. J. (1965). The effects of surgical exposures of dental pulps in germ-free and conventional laboratory rats. Oral Surg Oral Med Oral Pathol 20, 340-349. Kerekes, K. & Olsen, I. (1990). Similarities in the microfloras of root canals and deep periodontal pockets. Endod Dent Traumatol 6, 1-5. Kroes, I., Lepp, P. W. & Relman, D.A. (1999). Bacterial diversity within the human subgingival crevice. Proc Natl Acad Sci U S A 96, 14547–14552.

13 Kuipers J. G., Nietfeld, L., Dreses-Werringloer, U., Koehler, L., Wollenhaupt, J., Zeidler, H. & Hammer, M. (1999). Optimised sample preparation of synovial fluid for detection of Chlamydia trachomatis DNA by polymerase chain reaction. Ann Rheum Dis 58, 103-108. Kumar, S., Tamura, K., Jakobsen, I. B. & Nei, M. (2001). MEGA2: molecular evolutionary genetics analysis software. Bioinformatics 17, 1244-1245. Lane, D. J. (1991). 16S/ 23S rRNA sequencing. In Nucleic Acid Techniques in Bacterial Systematics, pp. 115–175. Edited by Stackebrandt, E. and Goodfellow, M. Chichester, UK: John Wiley and Sons. Le Goff, A., Bunetel, L., Mouton, C. & Bonnaure-Mallet, M. (1997). Evaluation of root canal bacteria and their antimicrobial susceptibility in teeth with necrotic pulp. Oral Microbiol Immunol 12, 318-322. Melen, I., Lindahl, L., Andreasson, L. & Rundcrantz, H. (1986). Chronic maxillary sinusitis. Definition, diagnosis and relation to dental infections and nasal polyposis. Acta Otolaryngol 101, 320-327. Molander, A., Reit, C., Dahlén, G. & Kvist, T. (1998). Microbiological status of root-filled teeth with apical periodontitis. Int Endod J 31, 1-7. Munson, M. A., Pitt-Ford, T., Chong, B., Weightman, A. & Wade, W. G. (2002). Molecular and cultural analysis of the microflora associated with endodontic infections. J Dent Res 81, 761-766. Nakazawa, F., Sato, M., Poco, S. E., Hashimura, T., Ikeda, T., Kalfas, S., Sundqvist, G. & Hoshino, E. (2000). Description of Mogibacterium pumilum gen. nov., sp. nov. and Mogibacterium vescum gen. nov., sp. nov., and reclassification of Eubacterium timidum (Holdeman et al. 1980) as Mogibacterium timidum gen. nov., comb. nov. Int J Syst Evol Microbiol 50, 679-688. Ngeow, W. C. (1999). Orbital cellulitis as a sole symptom of odontogenic infection. Singapore Med J 40, 101-103. Paju, S., Bernstein, J. M., Haase, E. M. & Scannapieco, F. A. (2003). Molecular analysis of bacterial flora associated with chronically inflamed maxillary sinuses. J Med Microbiol 52, 591- 597. Paster, B. J., Boches, S. K. , Galvin, J. L., Ericson, R. E., Lau, C. N., Levanos, V. A., Sahasrabudhe, A. & Dewhirst, F. E. (2001). Bacterial diversity in human subgingival plaque. J Bacteriol 183, 3770-3783. Paster, B. J., Falkler, W. A., Enwonwu, C. O. & 7 other authors (2002). Prevalent bacterial species and novel phylotypes in advanced noma lesions. J Clin Microbiol 40, 2187-2191. Polz, M. F. & Cavanaugh, C. M. (1998). Bias in template-to-product ratios in multitemplate PCR. Appl Environ Microbiol 64, 3724-3730.

14 Rodrigues, J. L. M., Silva-Stenico, M. E., Gomes, J. E., Lopes, J. R. S. & Tsai, S. M. (2003). Detection and diversity assessment of Xylella fastidiosa in field-collected plant and insect samples by using 16S rRNA and gyrB sequences. Appl Environ Microbiol 69, 4249-4255. Rolph, H. J., Lennon, A., Riggio, M. P., Saunders, W. P., MacKenzie, D., Coldero, L. & Bagg, J. (2001). Molecular identification of microorganisms from endodontic infections. J Clin Microbiol 39, 3282-3289. Rousée J. M., Bermond, D., Piemont, Y., Tournoud, C., Heller, R., Kehrli, P., Harlay, M. L., Monteil, H. & Jaulhac, B. (2002). Dialister pneumosintes Associated with Human Brain Abscesses. J Clin Microbiol 40, 3871–3873 Rupf, S., Kannengiesser, S., Merte, K., Pfister, W., Sigusch, B. & Eschrich, K. (2000). Comparison of profiles of key periodontal pathogens in periodontium and endodontium. Endod Dent Traumatol 16, 269-275. Saitou, N. & Nei, M. (1987). The Neighbor-joining method: a new Method for reconstructing phylogenetic trees. Mol Biol Evol 4, 406-425. Sakamoto, M., Umeda, M., Ishikawa, I. & Benno Y. (2000). Comparison of the oral bacterial flora in saliva from a healthy subject and two periodontitis patients by sequence analysis of 16S rDNA libraries. Microbiol Immunol 44, 643-652. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. Savarrio, L., Mackenzie, D., Riggio, M., Saunders, W. P. & Bagg, J. (2005). Detection of bacteraemias during non-surgical root canal treatment. J Dent 33, 293-303. Siqueira Jr., J. F. & Rôças, I. N. (2002). Dialister pneumosintes can be a suspected endodontic pathogen. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 94, 494-498. Siqueira Jr., J. F. & Rôças I. N. (2004). Polymerase chain reaction-based analysis of microorganisms associated with failed endodontic treatment. Oral Surg Oral Med Oral Pat Oral Rad Endod 97, 85-94. Smith, G. L., Socransky, S. S. & Smith C. M. (1989). Rapid method for the purification of DNA from subgingival microorganisms. Oral Microbiol Immunol 4, 47-51. Stackebrandt, E. & Goebel, B. M. (1994). Taxonomic note: a place for DNA-DNA reassociation and 16S rRNA sequence analysis in the present species definition in bacteriology. Int J Syst Bacteriol 44, 846-849. Sundqvist, G. (1976). Bacteriological studies of necrotic dental pulps (Dissertation). Umea, SE: University of Umea. Sundqvist, G. (1992a). Associations between microbial species in dental root canal infections. Oral Microbiol Immunol 7, 257-262. Sundqvist, G. (1992b). Ecology of the root canal flora. J Endod 18, 427-430. 15 Suzuki, M. T. & Giovannoni, S. J. (1996). Bias caused by template annealing in the amplification of mixtures of 16S rRNA genes by PCR. Appl Environ Microbiol 62, 625-630. Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 4673-4680. Wade, W. G., Spratt, D. A., Dymock, D. & Weightman, A. J. (1997). Molecular detection of novel anaerobic species in dentoalveolar abscesses. Clin Infect Dis 25, 235-236. Walsh, L. J. (1997). Serious complications of endodontic infections: some cautionary tales. Aust Dent J 42, 156-159. Weisburg, W. G., Barns, S. M., Pelletier, D. A. & Lane, D. J. (1991). 16S ribosomal DNA amplification for phylogenetic study. J Bacteriol 173, 697-703.

16 Order Clostridiales K12 16S

Eubacteriaceae Peptostreptococcaceae Clostridiaceae (acession numbers in numbers (acession Escherichia coli Escherichia from unexposed endodontic endodontic from unexposed (D14154) (D14154) (AB037874) (AB037874) (AB037875) (AB037875) (AB021702)(AB021702) TT (Z36296) (Z36296) TT (U13041) (U13041) (D14150) (D14150) TT (AJ005636) (AJ005636) TT (AB021701) (AB021701) Clostridiales TT TT (values above 50% are shown next next are shown above 50% (values TT TT TT (D14139) (D14139) (Y16716) (Y16716) (X73452) (X73452) TT TT TT (AB023970) (AB023970) (X77836) (X77836) TT (AJ006962) (AJ006962) TT (Y18189)(Y18189) JCM 8140 8140 JCM JCM (AF167711) (AF167711) TT (Z36272) (Z36272) TT TT (AB037910) (AB037910) TT TT (U13039)(U13039) (Z49863)(Z49863) ATCCATCC 700923 700923 ATCCATCC 33093 33093 ATCC 33099 33099 ATCC ATCC ATCC 700924 700924 ATCC ATCC clone AF_F05 (AY821869) (AY821869) AF_F05 AF_F05 clone clone ATCC 700697 700697 ATCC ATCC TT (X77851)(X77851) ATCC 27337 27337 ATCC ATCC TT ATCCATCC 9321 9321 ATCCATCC 700079 700079 DSM 1645 1645 DSM DSM DSMDSM 92 92 ATCC 700696 700696 ATCC ATCC TT DSM 11416 11416 DSM DSM clone AF_H06 (AY821870) (AY821870) AF_H06 AF_H06 clone clone DSMDSM 5272 5272 (X76649) (X76649) strainstrain TO-931 TO-931 DSM 12816 12816 DSM DSM ATCC 35896 35896 ATCC ATCC WW 1471 1471 TT ATCCATCC 33089 33089 ATCCATCC 9650 9650 four novel phylotypes (X76650) (X76650) strainstrain SYR SYR DSM 794 794 DSM DSM (X71846)(X71846) clone AG_D03 (AY821867) (AY821867) AG_D03 AG_D03 clone clone TT TT typetype(L37585) (L37585) 25763 25763 ATCC ATCC A A 17 DSM 286 286 DSM DSM FilifactorFilifactor villosus villosus Streptococcaceae Streptococcaceae AnaerococcusAnaerococcus lactolyticus lactolyticus EubacteriumEubacterium nodatum nodatum DSMDSM 316 316 Mogibacterium Mogibacterium ATCC 35928 35928ATCC ATCC (L35515) (L35515) MogibacteriumMogibacterium timidum timidum he number of nucleotide substitutions per site. per substitutions of nucleotide he number DSMDSM 6159 6159 FilifactorFilifactor alocis alocis MogibacteriumMogibacterium vescum vescum EubacteriumEubacterium minutum minutum AnaerococcusAnaerococcus prevotii prevotii MogibacteriumMogibacterium diversum diversum ClostridalesClostridales MogibacteriumMogibacterium neglectum neglectum MogibacteriumMogibacterium pumilum pumilum clone AG_G04 (AY821868) (AY821868) AG_G04 AG_G04 clone clone ClostridiumClostridium hiranonis hiranonis 9393 ClostridiumClostridium formicaceticum formicaceticum 9797 EubacteriumEubacterium brachy brachy EubacteriumEubacterium infirmum infirmum NatronincolaNatronincola histidinovorans histidinovorans 9696 PeptostreptococcusPeptostreptococcus anaerobius anaerobius ClostridiumClostridium felsineum felsineum SporobacterSporobacter termitidis termitidis SarcinaSarcina ventriculi ventriculi 100 100 PapillibacterPapillibacter cinnamivorans cinnamivorans ClostridiumClostridium haemolyticum haemolyticum ClostridiumClostridium botulinum botulinum AnaerobacterAnaerobacter polyendosporus polyendosporus 100 100 SarcinaSarcina maxima maxima 100100 100100 generated over 100 tree replications 100 over generated ClostridiumClostridium aldrichii aldrichii Clostridiaceae Clostridiaceae 100100 100 100 AcetivibrioAcetivibrio cellulolyticus cellulolyticus 100100 9999 100100 100100 9999 100100 6969 100100 5151 6565 100 100 100100 100 100 5757 0.05 0.05 0.05 7878 9595 100 100 16S rDNA phylogenetic inter-relationships of inter-relationships rDNA phylogenetic 16S 8585

Figure 1. rDNA. Bootstrap confidence values rDNA. Bootstrap confidence t represents Scale bar nodes). to the branching parenthesis). Distance matrix calculated by the Jukes and Cantor method (1969). Phylogenetic tree Phylogenetic (1969). Cantor method and Jukes by the calculated matrix Distance parenthesis). and rooted for & Nei (1987) of Saitou method the neighbor-joining by constructed infections, indicated in bold type, and reference strains from the order order from the strains type, and reference in bold indicated infections, Table 1. Highest-scored BLAST search results from 46 taxonomic units detected in unexposed root canal infections by the sequence analysis of 16S rDNA clonal libraries and percentage distribution of phylotypes per subject.

Identity Phylotypes distribution per subject (%) BLAST search result (%) AA AB AC AD AE AF AG Bacteroidales oral clone MCE7_120 E3 98 4.2 Bacteroidales oral clone MCE7_164 E2b 99 2.9 Bacteroides-like sp. oral clone X083 98 1.4 Burkholderia fungorum strain NW-2 100 7.4 Burkholderia phenazinium 99 5.7 Burkholderia sp. CI6 98 4.2 Campylobacter gracilis 99 1.4 Catabacter hongkongensis 100 2.9 Clostridiales oral clone MCE3_9 E1 93 9.1 15.0 Clostridiales oral clone P4PB_122 P3 92 5.0 Desulfobulbus sp. oral clone R004 98 20.8 Dialister invisus 100 28.8 2.9 4.2 3.0 10.0 Dialister pneumosintes 99 5.0 Eubacteriaceae oral clone P2PB_46 P3 99 20.0 33.3 6.1 15.0 Eubacterium sp. oral clone BU061 99 2.7 Eubacterium sp. oral clone CK047 99 4.2 Eubacterium sp. oral clone DA014 94 6.1 Eubacterium sp. oral clone FX028 98 4.2 Eubacterium tardum SC87K 99 23.3 11.4 12.1 Eubacterium timidum 99 27.3 10.0 Filifactor alocis 99 1.4 22.2 12.5 20.0 Filifactor sp. oral clone BP1-58 99 1.4 Filifactor sp. oral clone BP1-88 99 22.2 Flexistipes sp. E3_33 oral isolate 99 2.9 Lachnospiraceae bacterium oral clone BP1-14 99 5.7 Lachnospiraceae oral clone MCE7_60 E1 99 2.9 Lachnospiraceae oral clone MCE9_173 E4 99 34.3 16.7 Lactobacillus panis DAF 1 98 12.5 Megasphaera sp. oral clone BB166 98 7.4 5.0 Megasphaera sp. oral clone MCE3_141 P1 99 11.0 3.7 Mogibacterium neglectum 93 3.0 Peptostreptococcus micros ATCC 33270 100 3.0 Peptostreptococcus anaerobius clone LK54 99 1.4 14.8 Peptostreptococcus sp. oral clone CK035 99 15.0 Pseudoramibacter alactolyticus strain 23263T 99 15.1 8.6 Selenomonas sp. oral clone JS031 98 3.7 Selenomonas-like sp. oral strain FNA3 98 7.4 Treponema socranskii subsp. 04 99 3.0 Treponema sp. 6:H:D15A-4 98 4.1 Uncultured actinobacterium clone APe2_64 98 12.1 Uncultured bacterium BH017 99 6.1 Uncultured bacterium inhufec A-37 99 6.1 Uncultured bacterium W090 99 2.7 3.0 Uncultured Eubacterium clone PUS9.170 99 2.7 11.1 Unidentified oral bacterium RP55-6 98 2.7 Veillonella sp. ADV 360.00 99 87.5 Total number of phylotypes 14 2 9 11 9 13 9

18 CAPÍTULO 2

Artigo submetido ao periódico Oral Microbiology and Immunology em 09/10/2007 (Anexo 3)

Real Time PCR quantitative detection of Porphyromonas gingivalis and Tannerella

forsythia in primary endodontic infections

Authors: Daniel Saito1, Cristiane Pereira Borges Saito1, Luiz Lehmann Coutinho2, Fabiana de

Souza Cannavan3, Siu Mui Tsai3, José Francisco Höfling1, Reginaldo Bruno Gonçalves1*

* Corresponding author

1. Department of Oral Diagnosis, Microbiology and Immunology Division, Piracicaba Dental

School, State University of Campinas, Av. Limeira 901, 13414-903, Piracicaba, São Paulo, Brazil,

2. Department of Animal Science, University of São Paulo, Piracicaba, São Paulo Brazil

3. Center of Nuclear Energy in Agriculture, University of São Paulo, Piracicaba, São Paulo, Brazil

Running title: Real Time PCR of endodontic bacteria

19 Abstract

Sensitive detection of bacteria is an essential requisite for understanding the etiology of root canal infections and consequent periapical lesions. Porphyromonas gingivalis and Tannerella forsythia are oral bacteria capable of triggering intensified immunological responses. Quantitative detection of these bacterial species in root canal infections could bring important insight into their clinical relevance.

A Real Time PCR assay was applied to determine the levels of P. gingivalis and T. forsythia in 34 cases of primary endodontic infections. Total genomic DNA was extracted from intraradicular samples, followed by species-specific DNA amplification with primers targeting the bspA surface antigen gene of T. forsythia and the Arg-gingipain (rgp) gene of P. gingivalis. Relative bacterial levels were determined according to standard curves adjusted to genome copy numbers, followed by numerical correction with the absolute mass of DNA from each sample. Overall, P. gingivalis, T. forsythia and a coexistance of both species were encountered in 24%, 56% and 18% of the subjects. Adjusted genome copy levels ranged from zero to 1.26 x 104, and from zero to 1.28 x 105, respectively. No significant associations among the abundances of the target bacteria, or the combination of both, and the presence of pain of endodontic origin could be observed.

T. forsythia was highly prevalent and numerous in the study subjects, whereas P. gingivalis was moderately frequent and less abundant. Despite previous reports associating T. forsythia and P. gingivalis with specific clinical parameters, they did not display any correlations with endodontic symptoms.

20 Introduction

Root canal bacteria are the primary etiological agents of pulpal and periapical diseases and their persistence after instrumentation is still among the major causes of failure in the endodontic therapy. Whilst the oral cavity is currently estimated to harbor more than 500 bacterial species (Paster et al. 2001), only a restricted parcel is able to thrive inside the root canal. De facto, the microbial structure commonly observed in endodontic infections is only a stationary sequel of a highly dynamic process of bacterial succession, guided by ecological interactions played upon microbial elements and environmental factors established inside the canal (Figdor and Sunqvist 2007). Over the years, it has become generally accepted that primary endodontic infections are, in essence, a mixed composition of facultative and strictly anaerobic bacteria and, in this aspect, combinations of bacteria are expected to arise (Sundqvist 1992). Truly, it has been shown that particular associations of endodontic species can observed in the intraradicular habitat, either in asymptomatic (Peters et al. 2002), or in symptomatic teeth (Gomes et al. 2004). Specific associations of residual bacteria have also shown to be more frequent in cases of persistent periapical lesions in monkeys, when compared to the monoinfected or aseptic counterparts (Fabricius et al. 2006).

Associations of P. gingivalis and T. forsythia are directly involved with risk factors of periodontal disease (Grossi et al. 1995) and, together with Treponema denticola, constitute the ‘red-complex’, a group consisting of Gram negative anaerobic bacteria implicated with adult and refractory periodontal disease (Socransky et al. 2002, 1998). P. gingivalis is probably the most widely investigated oral pathogen and possesses a plethora of pathogenic properties including fimbriae, proteinases, lipoteichoic acids, exopolisaccharides, and hemin-binding proteins (Holt et al. 1999). T. forsythia harbors a multi-functional surface and secreted protein responsible for many pathogenic activities, including bacterial coaggregation, fibronectin and fibrinogen binding, epitelial cell attachment and invasion, and proinflamatory cytokine and chemokine induction (Inagaki et al. 2006). Interbacterial binding and growth induction by P. gingivalis and T. forsythia cells have been observed in vitro (Yoneda et al. 2005, Yao et al. 1996). This synergistic interaction is also evidenced in the in vivo scenario, where coinoculation of both species can trigger heightened pathogenicity, with abscess formation and sepsis (Yoneda et al. 2001, Takemoto et al. 1997).

Previous reports have shown positive correlations between endodontic bacterial load and size of periradicular lesion and periapical abscesses (Figdor and Sundqvist 2007). In this sense, the application of quantitative techniques can provide a more comprehensive view on the clinical relevance of particular species of bacteria (Gomes et al. 2004, Jung et al. 2001). A substantial number of microbiological examinations on root canal infections has relied on cultivation and, 21 more recently, on strictly qualitative molecular identification techniques such as the conventional Polimerase Chain Reaction (PCR). While cultivation provides invaluable information on growth and phenotypic properties of viable cells, it still lacks the necessary sensitivity for the very fastidious or as-yet uncultivable bacterial species (Rolph et al. 2001). Likewise, end-point PCR, a highly sensitive molecular technique broadly used for microbial identification, does not offer reliable quantitative capabilities for today's standards (Bustin et al. 2000, Ferre 1992).

The Real Time PCR technology provides precise fluorescence-based quantification of target DNA levels by a cycle-to-cycle monitoring of amplification products (Heid et al. 1996). Since fluorescence acquisition is done at the initial log-linear phase of DNA amplification, it circumvents common quantification biases inherent to end-point PCR analysis (Ferre et al. 1992). Also, as reaction products are directly detected inside the test tube, there is no the need for post processing, increasing automation and minimizing potential analysis errors (Bustin 2000). SYBR Green I is an intercalating dye that binds to the minor groove of double-stranded DNA. It can be applied with virtually any set of PCR primers, offering improved versatility and simplicity of use (Bustin 2000). It has been proved to be a sensitive, rapid and accurate reporter molecule for nucleic acid quantification, with specificity and robustness comparable to those of other available chemistries (Maeda et al. 2003, Morrison et al. 1998).

In the present study, a SYBR Green I Real Time PCR assay was used to determine the relative levels of P. gingivalis and T. forsythia in 34 cases of primary endodontic infections, in an effort to bring light into their clinical relevance.

Material and Methods

Subjects. Working approval was granted by the Ethical Committee for Human Subjects of the Dental School of Piracicaba. All subjects had been referred for endodontic treatment at the Dental School of Piracicaba, and were selected for the presence of pulpal necrosis by clinical and radiographic evaluations. One tooth per subject was included in the study. Overall, thirty four patients, 14 males and 20 females, ranging from 15 to 61 years old (mean 34.24 ± SD 13.73) were analyzed. Subjects who had undergone antibiotic therapy within two months prior to collection were not included. According to the observed clinical parameters, patients were classified in two categories for further correlation with the detected bacterial levels: symptomatic (n= 24) and asymptomatic (n= 10). Symptomatic cases were considered those harboring teeth associated with tenderness to percussion, sensibility to palpation, spontaneous pain, or a combination of those features. All other cases were considered asymptomatic.

22 Sample collection. Samples were collected as described elsewhere (Saito et al. 2006). Briefly, the teeth to be sampled were isolated with rubber dam, cleaned and opened using a sterile procedure. Intraradicular samples were obtained by the introduction of a sterile #15 K-file and 4 consecutive sterile paper points inside the root canal. The file and the paper points were placed in a test tube containing 1 mL of TE buffer (10 mM TrisHCl, 1 mM EDTA, 0.1 mM NaCl, pH 8.0) and immediately transported to the laboratory.

Bacterial strains. The following reference strains were used as controls for conventional PCR and Real Time PCR assays: Aggregatibacter actinomycetemcomitans ATCC 29522, Bacteroides fragilis ATCC 25285, Bacteroides merdae M-36, Bacteroides vulgatus ATCC 8482, Escherichia coli ATCC 12795, Porphyromonas assacharolytica ATCC 25260, Porphyromonas circumdentaria ATCC 51356, Porphyromonas endodontalis ATCC 35406, Porphyromonas gingivalis ATCC 33277, Porphyromonas levii ATCC 29147, Porphyromonas salivosa NCTC 11632, Prevotella oulora ATCC 43324, Prevotella intermedia ATCC 25611, Prevotella nigrescens NCTC 9336, Pseudomonas aeruginosa ATCC 10145, Streptococcus salivarius ATCC 25975, Streptococcus sanguinis ATCC 10556, Streptococcus sobrinus ATCC 27607, Streptococcus mutans ATCC 25175 and Tannerella forsythia ATCC 43037.

DNA extraction. DNA from clinical samples and reference bacteria was extracted as described elsewhere (Saito et al. 2006), ressupended in 30 µL of TE buffer and immediately stored at – 20 ºC until further required.

Preparation of standard DNA: Standard DNA solutions corresponding to 107 genome copies of P. gingivalis and T. forsythia were prepared based on their respective genome sizes (2.34 Mb for P. gingivalis (Nelson et al. 2003) and 3.41 Mb for T. forsythia (Tanner & Izard 2000), and considering the mean weight of a nucleotide pair to be 1.023 x 10-12 ng (Dolezel et al. 2003). These solutions were subsequently used to obtain serial 10-fold dilutions down to 101 genome copies per reaction.

Design of PCR primers. Primers targeting the arginin-specific cystein-proteinase (Arg-gingipain or rgp) gene of P. gingivalis were previously published (Morillo et al. 2004) (table 1). Primers for the surface antigen bspA gene of T. forsythia were obtained based on the complete gene sequence acquired from the GenBank database (accession number AF054892). The Primer3 software was used for the design of candidate oligonucleotide sequences (http://frodo.wi.mit.edu). A final set of primers was chosen according to the lowest potential to form secondary structures, as determined by analysis with the Netprimer software (http://www.premierbiosoft.com/netprimer) (table 1).

23 Validation of markers by end-point PCR. Conventional PCR pilots were performed in total volumes of 25 µL with the following conditions: 1x PCR Buffer, 2 mM MgCl2, 0.2 mM dNTPs, 0.2 μM primers, 1.5 U Platinum Taq DNA polymerase (Invitrogen). The temperature profiles were: a) for P. gingivalis (rgp gene): initial denaturation at 96°C for 3 min; 36 cycles of denaturation at 94°C for 1 min, annealing at 60°C for 1 min and extension at 72°C for 1min; final extension at 72°C for 10 min; b) for T. forsythia (bspA gene): initial denaturation at 96°C for 3 min; 36 cycles of denaturation at 94°C for 1 min, annealing at 46°C for 1 min and extension at 72°C for 1min; final extension at 72°C for 10 min. Supplementary PCR assays with universal 16S rDNA primers (table 1) were performed to assess the presence of bacteria in the clinical samples, according to the conditions described elsewhere (Saito et al. 2006).

Optimization of the Real Time PCR assay. Reactions were conducted in a Roche LightCycler 1.0 System under the LightCycler 4.24 Run Software (Roche Applied Science). Standard DNA corresponding to 103 genome copies of P. gingivalis and T. forsythia were used as templates.

Optimization was carried out with various MgCl2 and primer concentrations, as stated by the manufacturer's instructions (Roche). Final reaction mixtures were performed in total volumes of 10 µL containing 2.0 µL DNA, 1.0 µL LightCycler FastStart DNA Master SYBR Green I (Roche),

4mM MgCl2 and 0.5 μM of each primer. The temperature profiles were: a) for P. gingivalis (rgp gene): hot-start denaturation at 95°C for 10 min and 48 cycles of denaturation at 94°C for 10 s, annealing at 60°C for 7 s, extension at 72°C for 7 s and fluorescence acquisition at 78°C for 3 s; b) for T. forsythia (bspA gene): hot-start denaturation at 95°C for 10 min and 60 cycles of denaturation at 95°C for 10 s, annealing at 46°C for 7 s, extension at 72°C for 7 s and fluorescence acquisition at 83°C for 3 s.

Real Time PCR amplification of standard DNA. Amplification profiles for P. gingivalis and T. forsythia were generated from duplicate reactions using 10-fold dilutions of standard DNA (figure 1). The computer-assisted second derivative maximum algorithm was used for crossing point (CP) inference of each DNA dilution. Standard curves were generated by linear regression analysis, and used as a basis for further quantification of target DNA from the clinical samples (figure 1).

Real Time PCR amplification of DNA from clinical samples. Prior to quantification of target bacteria, the total ammount of DNA from each clinical sample was assessed by spectrophotometry at 260nm. Real Time PCR amplification assays were conducted in duplicates, and as described for standard DNA. Genome copy levels were inferred based on the previously obtained standard curves. In order to minimize potential variations in sample volume, absolute genome copy levels were normalized by numerical division by the total ammount of DNA measured from each sample. 3.0 µL of each Real Time PCR product was submited to 1.0% 24 agarose gel electrophoresis and stained with ethidium bromide to verify reaction efficiency and specificity.

Melting curve analysis. Reaction specificities were further verified by melting curve analysis (figure 2). The following temperature profiles were used: a) for P. gingivalis: a progressive temperature increase from 70°C to 95°C at a 0.1°C/s transition rate, with continuous fluorescence acquisition; b) for T. forsythia: a progressive temperature increase from 70°C to 95°C at a 0.1°C/s transition rate, with continuous fluorescence acquisition. To minimize potential primer-dimer interferences, fluorescence acquisition temperatures were adjusted to approximately 4°C below the melting point of each amplification product.

Sequencing of selected PCR products. Representative electrophoretic gel bands were selected to ascertain Real Time PCR amplicon identities from the clinical samples. DNA was purified with the GFX DNA Purification kit, as stated by the manufacturer (Amersham Biosciences). DNA sequencing reactions were performed in 10 μL mixtures containing approximatelly 100 ng template DNA, 1 μL Big Dye Terminator Ready version 3.0 (Applied Biosystems), 0.5 mM primer and 3 μL sequencing buffer (200 mM Tris/ HCl pH 9.0, 5 mM MgCl2). Nucleotide sequences were visualized and edited with the Bioedit 7.0.5.3 software (http://www.mbio.ncsu.edu/BioEdit) and submitted to the BLASTN algorithm (BLAST 2.0; http://www.ncbi.nlm.nih.gov/blast) for comparison with sequences from the GenBank database. A 98% nucleotide identity threshold was used for positive identification to species level (Saito et al. 2006).

Statistical analysis. The R (www.r-project.org) and BioEstat (http://www.mamiraua.org.br/) softwares were used for statistical analysis. Data exploration with the Lilliefors normality test indicated that data were non-homogeneous even after log transformation. The null hypothesis that the target species were not associated with endodontic symptoms was evaluated with the Mann- Whitney non-parametrical test, with a significance level α= 0.05. In order to assess the correlation between the coexistence of the target bacteria and the clinical parameters, the integrated levels of P. gingivalis and T. forsythia were used in the cases where both appeared concurrently; in all other cases (i.e. exclusive presence of either species or absence of both), the integrated levels were assigned as null. Lastly, potential interbacterial associations were assessed with the Fisher’s exact test.

Target species Orientation Position Oligonucleotide sequence (5’ – 3’) Amplicon size Reference

forward 1308 CCTACGTGTACGGACAGAGCTATA P. gingivalis 72 bp Morillo et al. 2004 (rgp gene) reverse 1379 AGGATCGCTCAGCGTAGCATT

T. forsythia forward 1911 TCACTATTGTGTCTCGCTG 133 bp this study (bspA gene)

25 reverse 2043 TCTCTCCGATTGTGGTTA

forward 8 AGAGTTTGATCCTGGCTCAG Total bacteria (fD1) 1534 bp Weisburg et al. 1991 (16S rRNA gene) reverse (rD1) 1541 AAGGAGGTGATCCAGCC Table 1. PCR primers used in this study.

Results

Real Time PCR amplification of standard DNA. No evidence of non-specific or cross-reaction products was observed for the bspA and rgp markers, as determined by conventional end-point PCR (data not shown). Melting curve analysis of Real Time PCR amplicons showed well-depicted peaks for both markers (figure 2). The minimum number of cycles required to amplify the full range of serially diluted DNA ranged from 23 to 43 for rgp, and from 24 to 54 for bspA (figure 1). The application of an amplification efficiency formula (Pfaffl 2001) revealed values of 2.0 and 1.6 for P. gingivalis and T. forsythia, provided that a 2.0 value corresponds to an optimum efficiency. Linear regression equations used for inference of genome copy levels of the target species in the clinical samples are shown in figure 1, along with the respective r2 values.

Real Time PCR amplification of DNA from clinical samples. All samples were positive for the presence of bacteria, as observed by conventional end-point PCR with 16S rDNA universal primers (data not shown). Overall, Real Time PCR detected P. gingivalis, T. forsythia and a coexistence of both species in 8/34 (24%), 19/34 (56%) and 6/34 (18%) of the subjects, respectively. Normalized genome copy levels in the clinical samples ranged from zero to 1.26 x 104, for P. gingivalis, and from zero to 1.28 x 105, for T. forsythia (mean 4.86 x 102 and 7.71 x 103 genome copies per sample, respectively).

Statistical analysis. No statistical difference between the symptomatic and asymptomatic groups was observed with the Mann-Whitney test, based on the normalized genome copy levels of either species [(p= 0.07 (P. gingivalis) and p= 0.46 (T. forsythia)] or the combination of both (p= 0.13). According to the Fisher’s exact test, no statistical significant conclusions could be drawn in regard to potential interbacterial associations among the target bacteria (p= 0.11).

26 Discussion

Real Time PCR is a highly sensitive technique that brings unprecedented contribution to pathogen quantification in mixed oral infections (Boutaga et al. 2006). Previous reports have shown that it can harbor 41-fold greater sensitivities than colony counting for oral anaerobic bacteria, offering detection increases of up to 36% and 51%, when considering the particular cases of T. forsythia and P. gingivalis (Verner et al. 2006, Martin et al. 2002). While such observations might be, on one hand, a direct consequence of the elevated cultivation demands of these fastidious species, on the other, they reflect the truly high accuracy and detection capabilities of the Real Time PCR technique. Indeed, it has been demonstrated that the technique can successfully differentiate a blank sample from another containing only 1 copy of target nucleic acid (Morrison et al. 1998).

In this study, a SYBR Green I – based Real Time PCR assay was applied to investigate the relative levels of the oral pathogens P. gingivalis and T. forsythia in primary endodontic infections. Because SYBR Green I will bind to any double-stranded DNA, a careful and throughout optimization step should be performed to avoid simultaneous quantification of potential non-specific products. We have taken preventive technical measures to minimize such biases, including preliminary in silico examination of primers, visualization of gel bands and analysis of melting curves. The evidence of well-defined unique peaks on melting curve analysis assured satisfactory reaction specificities for both markers, with no evidences of primer-dimer formation. As additional precautions, real time-monitoring of products was conducted by fluorescence acquisition at temperatures where double-stranded DNA was presumably composed only of specific products (Morrison et al. 1998) and representative bands from clinical samples were selected for DNA sequencing confirmation of amplicons.

The crossing-point (CP) is, by definition, the cycle number at which the fluorescence signal is acquired. It can be infered by a user-defined method, also known as the fit point, or by a computer-assisted algorithm, such as the second derivative. In this method, the CP corresponds to the cycle where the second derivative of the fluorescence intensity curve reaches its first maximum value and which, in turn, coincides with the beginning of the log-linear phase of the amplification curve (Luu-The et al. 2005). In our study, the second-derivative method was chosen due to its higher consistency and adequacy in the detection of low levels of target DNA, when compared to the fit point method (Luu-The et al. 2005).

Black-pigmented bacteria are among the most predominant organisms in oral anaerobic infections, commonly found in gingivitis, periodontitis, endodontic infections and dentoalveolar abscesses (van Winkelhoff et al. 1985). P. gingivalis, a Gram-negative rod, is probably the most

27 pathogenic and widely studied black-pigmented bacteria (Holt et al. 1999). Detection frequencies of P. gingivalis in root canal infections varies considerably. By the use of 16S rDNA PCR methodologies, Siqueira et al. (2004) and Fouad et al. (2002) were able to detect it in 4% of the samples, while Foschi et al. (2005) revealed the species in 13% and of the study cases. Higher prevalences were observed by Gomes et al. (2005) (38%) and Siqueira et al. (2002) (41%), also by PCR approaches. In this study, P. gingivalis was detected in 24% of the subjects, corroborating the results of a DNA-DNA hybridization assay of Siqueira et al. (2002) (28%) and an oligonucleotide probing study of Jung et al. (2001) (27%).

T. forsythia, formerly Bacteroides forsythus, is a Gram-negative anaerobic rod originally recovered from subjects with progressive advancing periodontitis, and afterwards isolated from various forms of periodontal disease, root canal infections and perimplantitis (Tanner and Izard 2006). The prevalence of T. forsythia in the root canal environment can also fluctuate considerably. While Rôças et al. (2001), Gomes et al. (2006), Fouad et al. (2002) and Jung et al. (2001) found the species in 26%, 24%, 21% and 18% of primary endodontic infections by PCR- based methodologies, Foschi et al. (2005) and Siqueira et al. (2004) were able to detect it only in 7% and 4% of the cases by conventional and nested 16S rDNA PCR, respectively. In the present study, T. forsythia was observed in 56% of the test subjects, an estimate close to those obtained by Siqueira and Rôças (2003) with nested PCR (52%) and by Gonçalves and Mouton (1999) with immunocapture-PCR (54%).

Relative disparities in the prevalence of intraradicular bacteria are usually expected when comparing the results from different authors. This might be due to a combination of factors, such as case selection criteria, sampling and DNA extraction methods, geographical origin of subjects, or detection techniques (Baumgartner et al. 2004, Siqueira et al. 2002). For instance, Munson et al. (2002) were unable to detect any clones of T. forsythia by 16S rDNA sequencing analysis, but revealed the presence of the species in one subject by cultivation. Similar high-throughput clone library studies were unable to detect either T. forsythia or P. gingivalis in primary or refractory endodontic infections (Saito et al. 2006, Rolph et al. 2001), indicating that detection of these particular species of bacteria could be dependent on the technique employed.

According to the corrected genome copy levels, P. gingivalis was not significantly associated with pain of endodontic origin, confirming the results of other molecular investigations (Foschi et al. 2005, Rôças et al. 2001 and Jung et al. 2000). Nonetheless, Hashioka et al. (1992) found representatives of Porphyromonas spp. to be implicated, as a group, with subacute endodontic symptoms in a cultivation study. We found no significant correlation between the detected levels of T. forsythia and endodontic symptoms, in agreement with the observations obtained by non-quantitative molecular investigations (Foschi et al. 2005, Siqueira et al. 2003, 28 Rôças et al. 2001 and Jung et al. 2000). A nested-PCR study conducted by Gomes et al. (2006), however, detected significant association between T. forsythia and tenderness to percussion in a larger sample set containing primary and treated cases of endodontic infections. Lastly, when considering P. gingivalis and T. forsythia as a complex, no indications of their involvement with clinical symptoms could be statistically evidenced in the present study.

In conclusion, the application of a Real Time PCR methodology has shown that T. forsythia can be highly prevalent and numerous in endodontic infections, whereas P. gingivalis is only moderately frequent and less abundant, displaying 16-fold lower average levels than those observed for T. forsythia. Our results reveal that P. gingivalis and T. forsythia do not play a primary role in the development of endodontic pain, either individually or in conjunction.

Acknowledgements

This work was supported by grants from the Brazilian funding agency Fundação de Amparo à Pesquisa do Estado de São Paulo – FAPESP (04/01674-6, 04/13548-5).

References

Baumgartner JC, Siqueira JF Jr, Xia T, Rocas IN. Geographical differences in bacteria detected in endodontic infections using polymerase chain reaction. J Endod 2004: 30: 141-144. Boutaga K, van Winkelhoff AJ, Vandenbroucke-Grauls CM, Savelkoul PH. The additional value of real-time PCR in the quantitative detection of periodontal pathogens. J Clin Periodontol 2006: 33: 427-433. Bustin SA. Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J Mol Endocrinol 2000: 25: 169–193. Dolezel J, Bartos J, Voglmayr H, Greilhuber J. Nuclear DNA content and genome size of trout and human. Cytometry 2001: 43: 2-11. Fabricius L, Dahlén G, Sundqvist G, Happonen R-P, Möller ÅJR. Influence of residual bacteria on periapical tissue healing after chemomechanical treatment and root filling of experimentally infected monkey teeth. Eur J Oral Sci 2006: 114: 278–285. Ferre F. Quantitative or semi-quantitative PCR: reality versus myth. PCR Methods Appl 1992: 2: 1-9. Figdor D, Sundqvist G. A big role for the very small--understanding the endodontic microbial flora. Aust Dent J 2007: 52: 38-51.

29 Foschi F, Cavrini F, Montebugnoli L, Stashenko P, Sambri V, Prati C. Detection of bacteria in endodontic samples by polymerase chain reaction assays and association with defined clinical signs in Italian patients. Oral Microbiol Immunol 2005: 20: 289-295. Fouad AF, Barry J, Caimano M, Clawson M, Zhu Q, Carver R, et al. PCR-based identification of bacteria associated with endodontic infections. J Clin Microbiol 2002: 40: 3223-3231. Gomes BPFA, Jacinto RC, Pinheiro ET, Sousa ELR, Zaia AA, Ferraz CCR, Souza-Filho FJ. Molecular analysis of Filifactor alocis, Tannerella forsythia, and Treponema denticola associated with primary endodontic infections and failed endodontic treatment. J Endod 2006: 32: 937-940. Gomes BPFA, Jacinto RC, Pinheiro ET, Sousa ELR, Zaia AA, Ferraz CCR, Souza-Filho FJ. Porphyromonas gingivalis, Porphyromonas endodontalis, Prevotella intermedia and Prevotella nigrescens in endodontic lesions detected by culture and by PCR. Oral Microbiol Immunol 2005: 20: 211-215. Gomes BPFA, Pinheiro ET, Gadê-Neto CR, Souza ELR, Ferraz CCR, Zaia AA, Teixeira FB, Souza-Filho FJ. Microbiological examination of infected dental root canals. Oral Microbiol Immunol 2004: 19: 71-76. Gonçalves RB, Mouton C. Molecular detection of Bacteroides forsythus in infected root canals. J Endod 1999: 25: 336-340. Grossi SG, Genco RJ, Machtei EE, Ho AW, Koch G, Dunford R, Zambon JJ, Hausmann E. Assessment of risk for periodontal disease. II. Risk indicators for alveolar bone loss. J Periodontol 1995: 66: 23-29. Hashioka K, Yamasaki M, Nakane A, Horiba N, Nakamura H. The relationship between clinical symptoms and anaerobic bacteria from infected root canals. J Endod 1992: 18: 558-561. Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitative PCR. Genome Res 1996: 6: 986-994. Holt SC, Kesavalu L, Walker S, Genco CA. Virulence factors of Porphyromonas gingivalis. Periodontol 2000. 1999: 20: 168-238. Inagaki S, Onishi S, Kuramitsu HK, Sharma A. Porphyromonas gingivalis vesicles enhance attachment, and the leucine-rich repeat BspA protein is required for invasion of epithelial cells by “Tannerella forsythia”. Infect Immun 2006: 5023–5028. Jung IY, Choi BK, Kum KY, Roh BD, Lee SJ, Lee CY, Park DS. Molecular epidemiology and association of putative pathogens in root canal infection. J Endod 2000: 26: 599-604. Jung IY, Choi BK, Kum KY, Yoo YJ, Yoon TC, Lee SJ, Lee CY. Identification of oral spirochetes at the species level and their association with other bacteria in endodontic infections. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2001: 92: 329-334.

30 Klein MI, Gonçalves RB. Detection of Tannerella forsythensis (Bacteroides forsythus) and Porphyromonas gingivalis by Polimerase Chain Reaction in subjects with different periodontal status. J Periodontol 2003: 74: 798-802. Luu-The V, Paquet N, Calvo E, Cumps J.Improved real-time RT-PCR method for high-throughput measurements using second derivative calculation and double correction. Biotechniques 2005: 38: 287-293. Maeda H, Fujimoto C, Haruki Y, Maeda T, Kokeguchi S, Petelin M, Arai H, Tanimoto I, Nishimura F, Takashiba S. Quantitative real-time PCR using TaqMan and SYBR Green for Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, tetQ gene and total bacteria. FEMS Immunol Med Microbiol 2003: 39: 81-86. Martin FE, Nadkarni MA, Jacques NA, Hunter N. Quantitative microbiological study of human carious dentine by culture and real-time PCR: association of anaerobes with histopathological changes in chronic pulpitis. J Clin Microbiol 2002: 40: 1698-1704. Morillo JM, Lau L, Sanz M, Herrera D, Mart_n C, Silva A. Quantitative real-time PCR based on single copy gene sequence for detection of periodontal pathogens. J Clin Peridontol 2004: 31: 1054–1060. Morrison TB, Weis JJ, Wittwer CT. Quantification of low-copy transcripts by continuous SYBR Green I monitoring during amplification. Biotechniques 1998: 24: 954-958, 960, 962. Munson MA, Pitt-Ford T, Chong B, Weightman A, Wade WG. Molecular and cultural analysis of the microflora associated with endodontic infections. J Dent Res 2002: 81: 761-766. Nelson KE, Fleischmann RD, DeBoy RT, Paulsen IT, Fouts DE, Eisen JA, Daugherty SC, Dodson RJ, Durkin AS, Gwinn M, Haft DH, Kolonay JF, Nelson WC, Mason T, Tallon L, Gray J, Granger D, Tettelin H, Dong H, Galvin JL, Duncan MJ, Dewhirst FE, Fraser CM. Complete genome sequence of the oral pathogenic bacterium Porphyromonas gingivalis strain W83. J Bacteriol 2003: 185: 5591-5601. Paster BJ, Boches SK, Galvin JL, Ericson RE, Lau CN, Levanos VA, Sahasrabudhe A, Dewhirst FE. Bacterial diversity in human subgingival plaque. J Bacteriol 2001: 183: 3770–3783. Peters LB, Wesselink PR, vanWinkelhoff AJ. Combinations of bacterial species in endodontic infections. Int Endod J 2002: 35: 698-702. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001: 29: 2002-2007. Rôças IN, Siqueira Jr. JF, Santos KRN, Coelho AMA. “Red complex” (Bacteroides forsythus, Porphyromonas gingivalis, and Treponema denticola) in endodontic infections: a molecular approach. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2001: 91: 468-471. Rolph HJ, Lennon A, Riggio MP, Saunders WP, MacKenzie D, Coldero L, Bagg J. Molecular identification of microorganisms from endodontic infections. J Clin Microbiol 2001: 39: 3282–3289. 31 Saito D, Leonardo RT, Rodrigues JLM, Tsai SM, Höfling JF, Gonçalves RB. Identification of bacteria in endodontic infections by sequence analysis of 16S rDNA clone libraries. J Med Microbiol 2006: 55: 101–107. Siqueira JF Jr, Rôças IN. Bacteroides forsythus in primary endodontic infections as detected by nested PCR. J Endod 2003: 29: 390-393. Siqueira JF Jr, Rôças IN, Alves FR, Santos KR. Selected endodontic pathogens in the apical third of infected root canals: a molecular investigation. J Endod 2004: 30: 638-643. Siqueira JF Jr, Rôças IN, De Uzeda M, Colombo AP, Santos KR. Comparison of 16S rDNA-based PCR and checkerboard DNA-DNA hybridisation for detection of selected endodontic pathogens. J Med Microbiol 2002: 51: 1090-1096. Socransky SS, Haffajee AD, Cugini MA, Smith C, Kent Jr. RL. Microbial complexes in subgingival plaque. J Clin Periodontol 1998: 25: 134-144. Socransky SS, Smith C, Haffajee AD. Subgingival microbial profiles in refractory periodontal disease. J Clin Periodontol 2002: 29: 260–268. Sundqvist G. Associations between microbial species in dental root canal infections. Oral Microbiol Immunol 1992: 7: 257–262. Takemoto T, Kurihara H, Dahlen G. Characterization of Bacteroides forsythus isolates. J Clin Microbiol 1997: 35:1378-1381. Tanner AC, Izard J. Tannerela forsythia, a periodontal pathogen entering the genomic era. Periodontol 2000 2006: 42: 88-113. van Winkelhoff AJ, Carlee AW, De Graff J. Bacteroides endodontalis and other black-pigmented Bacteroides species in odontogenic abscesses. Infect Immun 1985: 49: 494-498. Verner C, Lemaitre P, Daniel A, Giumelli B, Lakhssassi N, Sixou M. Carpegen real-time polymerase chain reaction vs. anaerobic culture for periodontal pathogen identification. Oral Microbiol Immunol 2006: 21: 341-346. Weisburg WG, Barns SM, Pelletier DA, Lane DJ. 16S ribosomal DNA amplification for phylogenetic study. J Bacteriol 1991: 173: 697–703. Yoneda M, Hirofuji T, Anan H, Matsumoto A, Hamachi T, Nakayama K, Maeda K. Mixed infection of Porphyromonas gingivalis and Bacteroides forsythus in a murine abscess model: involvement of gingipains in a synergistic effect. J Periodontal Res 2001: 36: 237-243. Yoneda M, Yoshikane T, Motooka N, Yamada K, Hisama K, Naito T, Okada I, Yoshinaga M, Hidaka K, Imaizumi K, Maeda K, Hirofuji T. Stimulation of growth of Porphyromonas gingivalis by cell extracts from Tannerella forsythia. J Periodontal Res 2005: 40: 105-109. Yao ES, Lamont RJ, Leu SP, Weinberg A. Interbacterial binding among strains of pathogenic and commensal oral bacterial species. Oral Microbiol Immunol 1996: 11: 35-41.

32

Figure 1. Amplification curves for P. gingivalis (A) and T. forsythia (B). Reactions performed with 10-fold serial dilutions of standard DNA ranging from 107 to 101 genome copies per reaction (left to right). Respective log-based standards curves (C and D) were determined by regression analysis using the crossing point (CP) values from duplicate reactions.

33 Figure 2. Melting profiles of Real Time PCR amplified standard DNA. The expected melting temperatures for P. gingivalis (rgp) and T. forsythia (bspA) were 82.8 °C and 87.5 °C, respectively.

34 CAPÍTULO 3

Trabalho apresentado no 24º. Congresso Brasileiro de Microbiologia (Out/2007) (Anexo 4). Artigo em vias de submissão ao periódico Journal of Clinical Microbiology

Intraradicular bacterial composition by

Terminal Restriction Length Polymorphism (T-RFLP) analysis

Authors: Daniel Saito1, Terence Lee Marsh2, José Francisco Höfling1, Reginaldo Bruno

Gonçalves1*

* Corresponding author

1. Department of Oral Diagnosis, Microbiology and Immunology Division, Piracicaba Dental

School, State University of Campinas, Av. Limeira 901, 13414-903, Piracicaba, São Paulo, Brazil

2. Center for Microbial Ecology, Michigan State University, East Lansing, MI, USA

Running title: Intraradicular bacteria by T-RFLP

35 Introduction

Under physiologic conditions, the dental pulp is free of any residing microorganisms. Nonetheless, physical or chemical trauma or caries infection can eventually involve the pulp tissue, leading to necrosis and subsequent endodontic infection. The necrotic pulp provides a selective habitat for the establishment of a complex microbiota, predominantly composed of strictly anaerobic bacterial species (Sundqvist 1992a). Bacteria residing in the endodontic milieu are able to engender destructive inflammatory responses in the perirradicular tissues, which in turn can give rise to periapical lesions (Stashenko et al. 1992) and important systemic complications (Bate et al. 2000, Breebaart et al. 2002). Elimination of bacteria from inside the canal is, therefore, a crucial goal in endodontic therapy.

Albeit substantial efforts have been taken towards the throughout characterization of the endodontic microbiota (Saito et al. 2006, Gomes et al. 2004, Sundqvist 1992b, Munson et al. 2002, Rolph et al. 2001), the intraradicular ecosystem still harbors plenty microbial complexity to warrant further investigations by contemporary methodologies. Previous reports have demonstrated that particular bacterial species can be responsible for the development of endodontic related signs and symptoms such as spontaneous and pre-operative pain, swelling, odor, tenderness to percussion and purulent exudate (Foschi et al. 2005, Gomes et al. 2004, Jacinto et al. 2003, Peters et al. 2002, Fouad et al. 2002, Hashioka et al. 1992). In light of these observations, it is reasonable to assume that the root canal bacterial communities can also encompass, as a whole, an important role in the development of clinical features. Beyond doubt, the application of microbial ecology tools that allow for a broader view on community structure could bring important information to the study of the root canal microbiota.

Community structure analysis is a microbial characterization approach that relies on two fundamental variables: species richness (the number of species in the community) and species evenness (the quantity of each species) (Liu et al. 1997). Assessment of these variables is impaired in traditional culture-based techniques, since cultivation can shift the original microbial composition by imposing additional selective pressures (Liu et al. 1997). In addition, approximately 50% of the oral bacteria cannot undergo cultivation (Paster et al. 2001, Aas et al. 2003) and, in this sense, community structure studies should be preferably conducted by molecular techniques capable of accounting the very fastidious and non-culturable taxa.

Terminal Restriction Length Polymorphism (T-RFLP) offers high-throughput quantitative analysis of community structure and community dynamics in complex environments (Blackwood et al. 2003, Osborn et al. 2000). In brief, the technique employs PCR targeting the bacterial 16S rRNA gene, in which one of the markers is fluorescently labeled at its 5’ end. The PCR product is

36 digested with an endonuclease with a 4-base pair recognition site, and the terminal restriction fragment is measured by an automated DNA sequencer (Marsh 2005, Marsh et al. 1999, Liu et al. 1997). Since different bacterial populations have distinct restriction recognition sites in the 16S rRNA gene, a genetic fingerprint of bacterial composition is obtained for each polymicrobial sample. T-RFLP has been successfully applied for comparing the bacterial composition of saliva from healthy and periodontal patients (Sakamoto et al. 2003), assessing shifts in microbial profiles after periodontal treatment (Sakamoto et al. 2004), investigating the effect of coronal restoration quality on endodontic bacterial composition (Hommez et al. 2004), and for comparing the bacterial communities of intraradicular samples and pus aspirates from symptomatic and asymptomatic teeth (Sakamoto et al. 2006).

In this study, the T-RFLP technique was applied to assess the intraradicular bacterial composition associated with asymptomatic, tender, and symptomatic endodontic infections, and to contrast the community structures associated with these three clinical categories.

Material and Methods

Subjects. Twenty five patients, 12 males and 13 females, ranging from 15 to 61 years old (mean 32.4± 13.7) were analyzed. Working approval was granted by the Ethical Committee for Human Subjects of the Piracicaba Dental School, State University of Campinas and written consent was required from all participants. Subjects harbored teeth with endodontic infections, and had been previously referred for endodontic treatment at the Piracicaba Dental School. Inclusion criteria included single or multi-rooted teeth with pulpal necrosis, with no distinction to the cause of endodontic involvement (caries or trauma). Teeth were evaluated for the presence of pulpal necrosis by visual inspection, probing, percussion, and radiographic evaluation. Subjects who had undergone antibiotic therapy within two months prior to collection or those who had been previously subjected to root canal treatment were not included. Specimens were classified in three study groups: asymptomatic (Group I, n= 7), tender (Group II, n= 5) and symptomatic (Group III, n= 13). Group I was composed of asymptomatic teeth, Group II of teeth tender to vertical percussion, and Group III of teeth harboring spontaneous pain with or without associated swelling or purulent exudate.

Sample collection. Each patient was submitted to local anesthesia and the tooth was isolated with a rubber dam. The crown was cleaned to eliminate debris and dental plaque. Antisepsis of the crown and operation field was conducted with 2.5% sodium hypochlorite for 1 min, followed by inactivation with 5% sodium thiosulfate (Ng et al. 2003). Coronal access cavity was gained by high-speed bur irrigated with sterile saline solution. Intraradicular samples were obtained by the

37 introduction of a sterile #15 K-file followed by 4 consecutive sterile paper points inside the root canal (Saito et al. 2006). The file and the paper points were placed in a test tube containing 1 mL of TE buffer (10 mM TrisHCl, 1 mM EDTA, 0.1 mM NaCl, pH 8.0) and immediately transported to the laboratory.

DNA extraction. The DNA from the clinical samples and reference bacteria were extracted as described elsewhere (Saito et al. 2006) and ressupended in 30 L of TE buffer, being immediately stored at – 20 ºC until further analysis.

Primer set selection. The Primer Sequence Prevalence Analysis program (MICA3 - http://mica.ibest.uidaho.edu/primer.php) was used to verify the range of detection of all possible combinations of the primers 27F, 63F, 1389R, 1392R and 1492R against the 16S rDNA Ribosomal Database Project II Release 9.37, with 1 mismatch allowed within 10 bases from the 5' end of either primer. The primer set 63F (5'- CAGGCCTAACACATGCAAGTC -3') and 1389R (5'- ACGGGCGGTGTGTACAAG -3') revealed the highest number of positive matches, and therefore was chosen for subsequent T-RFLP analyses. The 63F primer was labelled with 6'- carboxyfluorescein (6-FAM) at its 5' end to allow fluorescent detection of the terminal restriction fragment.

Amplification of DNA. 100 µL reactions were performed with 60 ng DNA, 1x PCR Buffer, 2 mM

MgCl2, 0.2 mM dNTPs (Promega), 0.4 µM primers (IDT Technologies) and 1.0 U Taq DNA polimerase (Invitrogen Co.), according to the following temperature profile: initial denaturation at 94 °C for 3 min; 25 cycles of denaturation at 94 °C for 1 min, annealing at 56°C for 1 min and extension at 72 °C for 2min; final extension at 72 °C for 10 min. 3 µL aliquots of each PCR product was submitted to 1.0% agarose gel electrophoresis and stained with ethidium bromide. PCR products were purified with the QIAquick PCR Purification Kit (Qiagen), as stated by the manufacturer.

Endonuclease restriction of amplified DNA. 200 ng of each purified product were restricted with HhaI (GCG’C) and MspI (C’CGG) endonucleases (New England Biolabs Inc.) in separate 15 µL reactions as instructed by the manufacturer. Aliquots of the original PCR products were left undigested and used as negative controls for T-RFLP analysis.

Generation of T-RFLP profiles: 2 uL of each restricted PCR product and the respective negative control were mixed with 7 uL deionized formamide and 1 uL MM1000 ROX size standard (Bioventures Inc.), denatured at 95 °C and immediately placed on ice. Duplicate mixtures were injected for 30 s into an ABI 3100 Sequence Analyzer (Applied Biosystems - Hitachi). T-RLFP profiles were generated by the Local Southern Method in GeneScan 3.7 Software (Applied Biosystems), using a 5 fluorescence units threshold.

38 Filtering and binning of T-RFLP profiles. T-RFLP data from duplicates were exported as spreadsheets, filtered by the IBEST analysis tools (Abdo et al. 2006) using a cutoff value of 6 standard deviations of total peak area, and binned (aligned) by the T-Align software (Smith et al. 2005) with a 1.0 base width. T-Align was also used to generate consensus profiles based on the average fluorescence intensities of corresponding peaks from the duplicates. Peaks that appeared exclusively in one of the duplicates were not included in the final consensus profiles. Peak height data were transformed into relative peak heights, as a ways to level up total fluorescence among samples after the removal of analytical noise.

Data analysis. Total, unique, single and double T-RFs were determined for each endonuclease with the EstimateS 8.0 software (Colwell 2006). The hypothesis that the T-RF richness among study groups was not statistically different was assessed with the T test with a significance α= 0.05. A Bray-Curtis dissimilarity matrix from the combined (concatenated) HhaI and MspI relative peak height data sets was obtained with the Vegan package for R (www.r-project.org) and used for cluster analysis with the MEGA 4 software (Tamura et al. 2007) according to the Unweighted Pair Group Method with Arithmetic mean (UPGMA). Principal Coordinates Analysis (PCoA) was accomplished based on the Bray-Curtis dissimilarity matrix from the previous step with the Canoco 4.5 software (Biometris). A quantitative display of relative peak heights from the consensus profiles of all study subjects was obtained with the Cluster 3.0 and TreeView 1.1 softwares for microarray data analysis (Eisen et al. 1998). T-RFs present in more than 30% of the subjects were predicted by the TAP-T-RFLP program (Marsh et al. 2000) based on a 16S rDNA datafile from most oral bacterial species (TRFMA; Nakano et al. 2006), adopting an error interval of 1 base pair. When no particular taxa could be predicted with this datafile, the 16S rDNA Ribosomal Database Project II Release 9.47 (http://rdp.cme.msu.edu) containing 16S rDNA sequences from 63,638 bacteria was used instead.

Results

All 25 samples were positive for bacteria, as evidenced by PCR with primers FAM-63F and 1389R. T-RFLP analysis with HhaI and MspI endonucleases revealed a total of 520 HhaI terminal restriction fragments (T-RFs(H)) and 501 MspI terminal restriction fragments (T-RFs(M)) in 25 cases of endodontic infections, with 123 unique T-RFs(H) and 122 unique T-RFs(M). T-RF richness per subject varied from 8 to 34 for HhaI, and from 8 to 32 for MspI analyses (table 1). No statistical significant differences of T-RF richness among the study groups was observed with either endonuclease, according to the T test (groups I and II: HhaI p= 0.12, MspI p= 0.57; groups I and III: HhaI p= 0.81, MspI p= 0.40; groups II and III: HhaI p= 0.19, MspI p= 0.96).

39 UPGMA analysis of combined HhaI and MspI profiles exhibited no evident tendency for clustering according to the predefined study groups (figure 1). Similarly, PCoA multivariate analysis did not reveal any distinct grouping of samples in broad, or as a function of the study groups (figure 2). The quantitative display (heatmap) of relative peak heights obtained with both endonucleases, along with TAP-T-RFLP prediction of the fragments found in more than 30% of the subjects, is displayed in figure 4.

HhaI and MspI endonucleases presented, in general, very similar resolving powers (table 1). Nonetheless, when comparing the results among study groups, slightly different behaviors could be observed: MspI revealed 11 (15%) more unique T-RFs than HhaI in Group I, while HhaI revealed 8 (13%) more unique T-RFs than MspI in Group II (table 1). The application of the T test demonstrated that such differences in enzyme efficiencies were not statistically significant, either within study groups (Group I p= 0.48, Group II p=0.26, Group II p= 0.76), or when considering the whole sample set (p= 0.67). With the exception of some highly predominant taxa, the observed bacterial assemblages were markedly variable. This is more easily conceived by the fact that 60/123 (48.8%) T-RF(H)s and 63/122 (51.6%) T-RF(M)s were detected in no more than 2 subjects (single plus double T-RFs) (table 1). In other words, when considering presence/absence data, half of the terminal fragments were very rare in our sample set and, therefore, could have accounted for most of the inter-subject variability observed. Highly predominant T-RFs that could be detected in more than half of the subjects included HhaI terminal restriction fragments T- RF(H)59 (Pedobacter sp. oral clone AV100, Flavobacterium-like sp. oral clone AZ123), T-RF(H)62 (Tannerella forsythia, Campylobacter spp.), T-RF(H)64 (Bacteroides spp.), T-RF(H)343 (Selenomonas sp. oral clone EZ011), T-RF(H)514 (Heliobacterium spp., Uncultured Chloroflexi bacterium) and T-RF(H)556 (Veillonella spp.), and MspI terminal restriction fragments T-RF(M)56 (Capnocytophaga spp., Dialister sp. oral clone BS095), T-RF(M)183 (Uncultured bacterium), T- RF(M)245 (Fusobacterium spp., Leptotrichia spp., Mogibacterium spp.), T-RF(M)258 (Selenomas sp. oral clone JI021), T-RF(M)266 (Peptococcus-like sp. oral clone I070, Actinomyces spp.), T- RF(M)267 (Selenomonas spp., Veillonella spp., Actinomyces spp.) and T-RF(M)464 (Peptostreptococcus sp. oral clone CK035).

Discussion

The intraradicular microbial communities have been extensively investigated by culture- based (Sundqvist 1992b, Gomes et al. 2004, Jacinto et al. 2003) and molecular-based techniques, including conventional PCR (Fouad et al. 2002, Conrads et al. 1997), Real Time PCR (Vianna et al. 2006), DNA-DNA checkerboard hybridization (Siqueira et al.), DGGE (Siqueira et al.

40 2004), and 16S rDNA clone libraries (Saito et al. 2006, Munson et al. 2002, Rolph et al. 2001). Terminal Restriction Length Polymorphism (T-RFLP) analysis brings important contribution to the study of the endodontic microbiota, allowing for rapid and efficient determination of community structure, also offering the possibility of taxonomical interpretation of terminal fragments (Marsh et al. 2000).

We have chosen primers 63F and 1389R for PCR amplification due to their higher range of detection when compared to other commonly used primers, as verified by computer-assisted preliminary evaluation (data not shown). Selection of this primer set was also favoured by the fact that the great majority of published 16S rDNA sequences are derived from the oligonucleotides 27F and 1492R; hence, th e use of primers that lay internally to such residues may facilitate the in silico prediction of T-RFs (Osborn et al. 2000). The election of HhaI and MspI as our restriction enzymes was founded on their improved efficiency in discriminating T-RFs on the basis of taxonomical information, when compared to other tetrameric endonucleases (Moyer at al. 1996).

An average of 20.8 T-RFs(H) and 20.0 T-RFs(M) per subject was obtained, indicating that, for the analysis of endodontic bacterial communities, both enzymes presented comparable restriction efficiencies. In regard to the taxonomical interpretation of the most prevalent fragments, however, HhaI has proved to be more proficient in resolving T-RFs related to Bacteroides spp. and Veillonella spp., while MspI was more capable of discriminating T-RFs representing Campylobacter spp. (figure 4). It is important to emphasize, however, that the number of terminal fragments identified in the present study does not reflect the actual diversity of the microbiota under investigation. In fact, fingerprinting techniques do not encompass sufficient sensitivity to detect the rare taxa, but still compose powerful and efficient tools for comparing biodiversity levels among complex ecosystems (Danovaro et al. 2007).

We are aware that T-RF prediction based on fragment length information may, at times, lack sufficient precision for single taxa inference (i.e. one T-RF may correspond to more than one taxonomic unit) (Nakano et al. 2006, Marsh 2005). Nonetheless, using a restricted database composed of 16S rDNA sequences from oral bacteria (TRFMA) reduced substantially the number of taxa correspondent to a particular fragment length and, consequently, increased the precision of our taxonomic inference. In reality, in silico prediction of the most predominant fragments revealed the presence of strict anaerobic bacteria from the genera Actinomyces, Bacteroides, Capnocytophaga, Eubacterium, Fusobacterium, Petptococcus, Peptostreptococcus, Selenomonas and Veillonella (figure 4), which have been extensively reported as common inhabitants of the root canal microenviroment (Saito et al. 2006, Gomes et al. 2004, Jacinto et al. 2003, Rolph et al. 2001, Sundqvist 1992b). 41 Terminal fragment T-RF(M)56 was one of the most prevalent in our study subjects (17/25), and was predicted as Tannerella forsythia, a Gram-negative filament-shaped anaerobic bacteria which bears a multi-functional protein responsible for a number of pathogenic properties (Holt et al. 2000), and has been considered as a putative periodontal pathogen due to its high prevalence in disease patients (Paster et al. 2001). A recent study by Gomes et al. (2006) revealed a significant association between T. forsythia and tenderness to percussion in primary and secondary endodontic infections. In our investigation, T. forsythia was evenly distributed among our study groups and did not display any evident trend towards symptoms. Other highly prevalent terminal fragments present in more than 17/25 (68%) of the subjects included T-RF(M)245 (Fusobacterium spp., Leptotrichia spp., Microbacterium sp. oral strain C24KA), T-RF(M)258 (Selenomonas sp. oral clone JI021), T-RF(H)343 (Selenomonas sp. oral clone EZ011) and T- RF(H)556 (Veillonella spp.).

Fragments predicted as Porphyromonas-like sp. oral clone DA064 (T-RF(H)64) and representatives of Bacteroides spp. (T-RF(M)58), were detected in 14/25 and 8/25 of the patients, respectively. Porphyromonas spp. and Bacteroides spp. are putative pathogenic bacteria in endodontic infections, significantly implicated with odor, spontaneous pain and tenderness to percussion (Gomes et al. 2004, Jacinto et al. 2003, Hashioka et al. 1992). Likewise, T-RFs corresponding to D. pneumosintes (T-RF(M)271) and Dialister oral clones BS095 (T-RF(H)558 and T-RF(M)52), BS016 and MCE7_134 (T-RF(M)271) were detected in more than 30% of the cases (figure 4). Dialister is a genus composed of anaerobic Gram-negative bacilli and has been progressively reported in both primary (Jacinto et al. 2007, Saito et al. 2006, Munson et al. 2002) and refractory (Rolph et al. 2001) endodontic infections by 16S rDNA analysis. Our results accord with previous evidence (Siqueira and Rôças 2002), in the sense that representatives of this genus can be highly frequent in the endodontic milieu.

Restriction fragments TRF(H)433 and TRF(H)460 were both detected in 11 subjects, out of which 10 either tender or symptomatic. TRF(H)433 was presumed as Campylobacter sp. oral clones BB120 and HB035, phylotypes originally reported in a case of refractory periodontitis (Paster et al. 2001) and on the hard palate of healthy subjects (Aas et al. 2005). TRF(H)460 was predicted as Bergeyella sp. oral clone AK152, a phylotype first reported by B. J. Paster and coworkers in an unpublished work on subgingival plaque (GenBank accession number AY008691), and shares a 99.7% nucleotide identity with Bergeyella sp. clone AF14, a recently detected phylotype from the amniotic fluid and subgingival plaque of a patient with preterm birth (Han et al. 2006). Bergeyella zoohelcum, the genus type species, is found in the oral cavity of dogs and cats and is frequently associated with bite wounds in humans (Reina and Borrell 1992,

42 Shuckla et al. 2004). Taking these evidences together, a pathological role for Bergeyella sp. in endodontic infections is conceivable.

In the present study, the weight of the intraradicular bacterial assemblages in the development of tenderness to percussion and spontaneous pain was investigated. Overall, we did not find any significant differences in T-RF richness or community structures among asymptomatic, tender and symptomatic endodontic infections. These observations were coherently achieved by univariate and multivariate statistical analyses and by hierarchical clustering. In this regard, our results discord from those of Sakamoto et al. (2006) and Siqueira et al. (2004), who detected differences in the bacterial composition of asymptomatic and symptomatic endodontic infections by T-RFLP and DGGE methodologies. However, both authors used specimens from different disease sites (root canals for asymptomatic infections and pus aspirates for symptomatic infections) and Sakamoto et al. (2006) used three additional restriction enzymes for T-RFLP analysis, possibly contributing to the observed variations in the bacterial profiles. In our investigation, samples were recovered from a single disease site (root canals). In this context, the inexistence of detectable differences in community structures among the study groups could also be a direct consequence of the more concise sampling strategy adopted herein.

Conclusions

T-RFLP analysis indicated that the bacterial composition of root canal infections is highly variable among individuals, with approximately 50% of the fragments present in no more than 2 of the subjects. Nonetheless, certain taxa displayed ubiquitous presence, namely Tannerella forsythia, Selenomonas sp. oral clones JI021 and EZ011, and Veillonella spp. The inexistence of statistically significant differences in T-RF richness among study groups, along with the lack of clearly visible clusters on hierarchical and multivariate analyses, supports the hypothesis that the intraradicular bacterial community assemblages, in entirety, do not play a role in the development of endodontic symptoms.

Acknowledgements

This work was supported by Coordenação de Aperfeiçoamento de Pessoal de Ensino Superior (CAPES; process BEX0567/06-0), a Brazilian federal government agency, and Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP; process 04/01674-6), a São Paulo state government agency. We thank Ms. Shari Tjugum-Holland from the Research Technology Support Facility (Michigan State University) for technical assistance with T-RFLP analyses.

43 References

Aas, J. A., Paster B. J., L. N. Stokes, I. Olsen, and F. E. Dewhirst. 2005. Defining the normal bacterial flora of the oral cavity. J. Clin. Microbiol. 43: 5721–5732. Abdo, Z, U. M. Schüette, S. J. Bent, C. J. Williams, L. J. Forney, and P. Joyce. 2006. Statistical methods for characterizing diversity of microbial communities by analysis of terminal restriction fragment length polymorphisms of 16S rRNA genes. Environ. Microbiol. 8: 929-938. Bate, A. L., J. K. Ma, and T. R. Pitt Ford. 2000. Detection of bacterial virulence genes associated with infective endocarditis in infected root canals. Int. Endod. J. 33: 194–203. Blackwood, C. B., D. Hudleston, D. R. Zak, and J. S. Buyer. 2007. Interpreting ecological diversity indices applied to terminal restriction fragment length polymorphism data: insights from simulated microbial communities. Appl. Environ. Microbiol.73: 5276-5783. Blackwood, C. B., T. Marsh, S. H. Kim, and E. A. Paul. 2003. Terminal restriction fragment length polymorphism data analysis for quantitative comparison of microbial communities. Appl. Environ. Microbiol. 69: 926-932. Breebaart, A. C., J. W. Bijlsma, and W. van Eden. 2002. 16-year remission of rheumatoid arthritis after unusually vigorous treatment of closed dental foci. Clin. Exp. Rheumatol. 20: 555–557. Colwell, R. K. EstimateS: statistical estimation of species richness and shared species from samples. Version 8.0. Persistent URL: http://purl.oclc.org/estimates. Conrads, G., S. E. Gharbia, K. Gulabivala, F. Lampert, and H. N. Shah. 1997. The use of a 16S rDNA directed PCR for the detection of endodontopathogenic bacteria. J. Endod. 23: 433-438. Danovaro, R., G. M. Luna, and A. Dell’Ano. 2007. Measuring species richness based on microbial community fingerprints: the emperor has no clothes. Appl. Environ. Microbiol. 73: 2399–2401. Eisen, M. B., P. T. Spellman, P. O. Brown, and D. Botstein. 1998. Cluster analysis and display of genome-wide expression patterns. Proc. Natl. Acad. Sci. U. S. A. 95: 14863–14868. Foschi, F., F. Cavrini, L. Montebugnoli, P. Stashenko, V. Sambri, and C. Prati. 2005. Detection of bacteria in endodontic samples by polymerase chain reaction assays and association with defined clinical signs in Italian patients. Oral Microbiol. Immunol. 20: 289–295. Fouad, A. F., J. Barry, M. Caimano, M. Clawson, Q. Zhu, R. Carver, K. Hazlett, and J. D. Radolf. 2002. PCR-Based identification of bacteria associated with endodontic infections. J. Clin. Microbiol. 40: 3223–3231. Gomes, B. F. P. A., E. T. Pinheiro, C. R. Gadê-Neto, E. L. R. Souza, C. C. R. Ferraz, A. A. Zaia, F. B. Teixeira, and F. J. Souza-Filho. 2004. Microbiological examinations of infected dental root canals. Oral Microbiol. Immunol. 19: 71-76. Han, Y. H., A. Ikegami, N. F. Bissada, M. Herbst, R. W. Redline, and G. G. Ashmead. 2006. Transmission of an uncultivated Bergeyella strain from the oral cavity to amniotic fluid in a case of

44 preterm birth. J. Clin. Microbiol. 44:1475–1483. Hashioka, K., M. Yamasaki, A. Nakane, N. Horiba, and H. Nakamura. 1992. The relationship between clinical symptoms and anaerobic bacteria from infected root canals. J. Endod. 18: 558- 561. Holt, S. C., L. Kesavalu, S. Walker, and C. A. Genco. 1999. Virulence factors of Porphyromonas gingivalis. Periodontol. 2000. 20: 168-238. Hommez, G. M. G., R. Verhelst, G. Claeys, M. Vaneechoutte, and R. J. G. De Moor. 2004. Investigation of the effect of the coronal restoration quality on the composition of the root canal microflora in teeth with apical periodontitis by means of T-RFLP analysis. Int. Endod. J. 37: 819– 827. Jacinto, R. C., B. P. F. A. Gomes, M. Desai, D. Rajendram, H.N. Shah. 2007. Bacterial examination of endodontic infections by clonal analysis in concert with denaturing high- performance liquid chromatography. Oral Microbiol. Immunol. 22: 403–410. Jacinto, R. C., B. F. P. A. Gomes, C. C. R. Ferraz, A. A. Zaia, and F. J. Souza-Filho. 2003. Microbiological analysis of infected root canals from symptomatic and asymptomatic teeth with periapical periodontitis and the antimicrobial susceptibility of some isolated bacteria. Oral Microbiol. Immunol. 18: 285-292. Liu, W. T., T. L. Marsh, H. Cheng, and L. J. Forney. 1997. Characterization of microbial diversity by determining terminal restriction fragment length polymorphisms of genes encoding 16s rRNA. Appl. Eviron. Microbiol. 63: 4516-4522. Marsh, T. L. 1999. Terminal restriction fragment length polymorphism (T-RFLP): an emerging method for characterizing diversity among homologous populations of amplification products. Curr. Opin. Microbiol. 2: 323-327. Marsh, T. L., P. Saxman, J. Cole, and J. Tiedje. 2000. Terminal restriction fragment length polymorphism analysis program, a web-based research tool for microbial community analysis. Appl Environ Microbiol. 66: 3616-3620. Marsh, T. L. 2005. Culture-independent microbial community analysis with terminal restriction fragment length polymorphism. Methods Enzymol. 397: 308-329. Moyer, C., J. M. Tiedje, F. C. Dobbs, and D. M. Karl. 1996. A computer-simulated restriction fragment length polymorphism analysis of bacterial small-subunit rRNA genes: efficacy of selected tetrameric restriction enzymes for studies of microbial diversity in nature. Applied Environ. Microbiol. 62: 2501–2507. Munson, M. A., T. Pitt-Ford, B. Chong, A. Weightman, and W. G. Wade. 2002. Molecular and cultural analysis of the microflora associated with endodontic infections. J. Dent. Res. 81: 761– 766. Nakano, Y., T. Takeshita, and Y. Yamashita. 2006. TRFMA: a web-based tool for terminal 45 restriction fragment length polymorphism analysis based on molecular weight. Bioinformatics. 22: 1788–1789. Ng Y. L., D. Spratt, S. Sriskantharajah, and K. Gulabivala. 2003. Evaluation of protocols for field decontamination before bacterial sampling of root canals for contemporary microbiology techniques. J. Endod. 29: 317-320. Osborn, A. M., E. R. Moore, and K. N. Timmis. 2000. An evaluation of terminal-restriction fragment length polymorphism (T-RFLP) analysis for the study of microbial community structure and dynamics. Environ. Microbiol. 2: 39-50. Paster, B. J., S. K. Boches, J. L. Galvin, R. E. Ericson, C. N. Lau, V. A. Levanos, A. Sahasrabudhe, and F. E. Dewhirst. 2001. Bacterial diversity in human subgingival plaque. J Bacteriol. 183: 3770–3783. Reina, J., and N. Borrell. 1992. Leg abscess caused by Weeksella zoohelcum following a dog bite. Clin. Infect. Dis. 14:1162–1163. Rolph, H. J., A. Lennon, M. P. Riggio, W. P. Saunders, D. MacKenzie, L. Coldero, and J. Bagg. 2001. Molecular identification of microorganisms from endodontic infections. J. Clin. Microbiol. 39: 3282–3289. Saito, D., R. T. Leonardo, J. L. M. Rodrigues, S. M. Tsai, J. F. Höfling, and R. B. Gonçalves. Identification of bacteria in endodontic infections by sequence analysis of 16S rDNA clone libraries. 2006. J. Med. Microbiol. 55: 101–107. Sakamoto, M., I. N. Rôças, J. F. Siqueira Jr., and Y. Benno. 2006. Molecular analysis of bacteria in asymptomatic and symptomatic endodontic infections. Oral Microbiol. Immunol.: 21: 112–122. Sakamoto, M., Y. Huang, M. Ohnishi, M. Umeda, I. Ishikawa, and Y. Benno. 2004. Changes in oral microbial profiles after periodontal treatment as determined by molecular analysis of 16S rRNA genes. J. Med. Microbiol. 53: 563–571. Sakamoto, M., Y. Takeuchi, M. Umeda, I. Ishikawa, and Y. Benno. 2003. Application of terminal RFLP analysis to characterize oral bacterial flora in saliva of healthy subjects and patients with periodontitis. J. Med. Microbiol: 52: 79–89. Shukla, S. K., D. L. Paustian, P. J. Stockwell, R. E. Morey, J. G. Jordan, P. N. Levett, D. N. Frank, and K. D. Reed. 2004. Isolation of a fastidious Bergeyella species associated with cellulitis after a cat bite and a phylogenetic comparison with Bergeyella zoohelcum strains. J. Clin. Microbiol. 42: 290–293. Siqueira, J. F. Jr, and I. N. Rôças. 2002. Dialister pneumosintes can be a suspected endodontic pathogen. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 94: 494-498. Siqueira, J. F. Jr, I. N. Rôças, and A. S. Rosado. 2004. Investigation of bacterial communities associated with asymptomatic and symptomatic endodontic infections by denaturing gradient gel electrophoresis fingerprinting approach. Oral Microbiol. Immunol. 19: 363–370. 46 Smith, C. J., B. S. Danilowicz, A. K. Clear, F. J. Costello, B. Wilson, and W. G. Meijer. 2005. T- Align, a web-based tool for comparison of multiple terminal restriction fragment length polymorphism profiles. FEMS Microbiol. Ecol. 54: 375-80. Stashenko, P., S. M. Yu, and C. Y. Wang. 1992. Kinetics of immune cell and bone resorptive responses to endodontic infections. J. Endod. 18: 422-426. Sundqvist, G. 1992a. Ecology of the root canal flora. J. Endod. 18: 427–430. Sundqvist, G. 1992b. Associations between microbial species in dental root canal infections. Oral Microbiol. Immunol. 7: 257-262. Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24: 1596-1599. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22: 4673–4680. Vianna, M. E., G. Conrads, B. P. F. A. Gomes, and H. P. Horz. 2006. Identification and quantification of archaea involved in primary endodontic infections. J. Clin. Microbiol. 44: 1274– 1282.

47 Table 1. Terminal restriction fragments identified in 25 cases of endodontic infections (identified as S1 – S25). HhaI MspI Group I. Asymptomatic (n= 7) S2 21 31 S8 8 21 S11 22 15 S15 18 13 S18 26 20 S19 18 27 S22 23 26 Total Group I 136 153 Distinct 73 84 Single* 42 51 Double** 14 15 Mean 19.4 21.9

Group II. Tender (n= 5) S5 28 12 S10 27 31 S14 25 25 S17 21 9 S25 21 19 Total Group II 122 96 Distinct 70 62 Single* 38 30 Double** 20 20 Mean 24.4 19.2

Group III. Symptomatic (n= 13) S1 11 8 S3 14 16 S4 27 25 S6 17 13 S7 18 23 S9 13 16 S12 21 20 S13 15 22 S16 34 32 S20 26 23 S21 21 20 S23 26 16 S24 19 18 Total Group III 262 252 Distinct 95 93 Single* 40 43 Double** 16 19 Mean 20.1 19.4 Overall Total 520 501 Overall Distinct 123 122 Overall Single* 35 42 Overall Double** 25 21 Overall Mean 20.8 20.0 * Exclusive to one subject; ** Exclusive to two subjects. 48

S5 Figure 1. UPGMA dendogram of T-RFLP S18 S4 profiles from 25 cases of endodontic S22 infections. A combined HhaI and MspI data S23 set was used to obtain a peak height S12 S24 sensitive Bray-Curtis dissimilarity matrix. S11 S21 Legend: S20 Asymptomatic ( ) S10 ◦ S16 Tender (•) S2 Symptomatic (•) S17 S14 S9 S6 S7 S19 S3 S8 S13 S25 S1 S15

0.4 0.3 0.2 0.1 0.0

Figure 3. Principal Coordinates Analysis plot (1st and 2nd axes) of T-RFLP profiles obtained from 25 cases of endodontic infections. A combined HhaI and MspI data set was used to obtain a peak height sensitive Bray-Curtis dissimilarity matrix.

Legend: Asymptomatic (◦) Tender (•) Symptomatic (•)

49

I II III : I : I Capnocytophaga spp. * Dialister sp. oral clone BS095

Tannerella forsythia *

Bacteroides spp.* Flavobacterium-like sp. oral clone AZ105 Porphyromonas-like sp. oral clone DA064

Eubacteriaceae oral clone MCE10_174 * Eubacterium sp. oral clones JS001/ FX033 Lachnospiraceae oral clone MCE9_173 Leptotrichia sp. oral clone EI013 * Desulfomicrobium orale

Uncultured bacterium †

Fusobacterium spp.* Leptotrichia-like sp. oral clone BB002 Leptotrichia sp. oral clone FB074 Microbacterium sp. oral strain C24KA I endonucleases from 25 cases of of from 25 cases I endonucleases Selenomonas sp. oral clone JI021 * I Peptococcus-like sp. oral clone I070 * Actinomyces sp. oral clones AP064/ BL008 Msp Dialister sp. oral clones BS016/ MCE7_134 * Dialister pneumosintes I and Msp I and

Actinomyces spp.*

Selenomonas spp. Hha Veillonella spp.

Campylobacter sp. oral clones BB120/ HB035 *

Firmicutes oral clone MCE7_107 * Catonella sp. oral clone AH153 * Eubacterium sp. oral strains A35MT/ BU061 Eubacteriaceae oral clone P2PB_46 P3 Bergeyella sp. oral clone AK152 * Peptostreptococcus sp. oral clone CK035 * * Capnocytophaga sp. oral clone ID062 Capnocytophaga gingivalis Uncultured Actinobacterium † Uncultured Chloroflexi bacterium * Capnocytophaga sputigena Sphingobacterium multivorum † Uncultured bacterium * Pedobacter sp. oral clone AV100 Flavobacterium-like sp. oral clone AZ123

* Tannerella forsythia Campylobacter sp. oral clones BB120/ HB035

* Bacteroides spp. aks are shown. Subjects are divided according to the observed parameters clinical according to the observed divided are aks Subjects are shown.

* Fusobacterium spp. Leptotrichia-like sp. oral clone BB002 Leptotrichia sp. oral clone FB074 Actinomyces sp. oral clone IO077 Kingella sp. oral clone ID059

* Actinomyces sp. oral clone AG004 Firmicutes oral clone CK057 I * Selenomonas sp. oral clone EZ011 (heatmap) of relative T-RF peak heights obtained with obtained of heights (heatmap) relative T-RF peak

Hha

† Heliobacterium sulfidophilum Heliobacterium gestii Uncultured Chloroflexi bacterium

* Burkholderia sp. oral clone AK168 * Selenomonas sputigena

* Veillonella spp. Quantitative display Quantitative

* Peptococcus sp. oral clone MCE10_265 Dialister sp. oral clone BS095

† Uncultured Gamma Proteobacterium Fragment predicted with the TRFMA Oral database Oral database with the TRFMA predicted Fragment endodontic infections. TAP-T-RFLP predicted pe predicted TAP-T-RFLP infections. endodontic Figure 4. Figure 4. (asymptomatic), II (tender), III (symptomatic). III (symptomatic). II (tender), (asymptomatic), * database 9.47 the RDP Release with predicted † Fragment

50 CONSIDERAÇÕES GERAIS

Na presente tese, a microbiota endodôntica foi avaliada por meio de três técnicas independentes de cultivo laboratorial: análise de bibliotecas clonais 16S rDNA (Fase I), PCR em Tempo Real (Fase II) e T-RFLP (Fase III). As técnicas foram aplicadas em amostras microbiológicas intra-radiculares de naturezas diversas: a análise de bibliotecas clonais foi realizada em amostras assintomáticas não-expostas ao meio oral; PCR em Tempo Real foi aplicada em amostras assintomáticas e sintomáticas; e T-RFLP foi utilizada em amostras assintomáticas, sensíveis à percussão e sintomáticas. Devido aos detalhes técnicos inerentes às diferentes metodologias e grupos de estudo utilizados, o universo amostral das três fases variou consideravelmente: 7 amostras para a Fase I, 34 para a Fase II e 25 para a Fase III.

Em geral, a análise de bibliotecas clonais demonstrou uma predominância de espécies anaeróbias, especialmente as Gram-positivas do filo Firmicutes. Apesar da relativa homogeneidade deste grupo de estudo, uma alta variação na composição bacteriana pôde ser observada entre os pacientes avaliados, constatada principalmente pela alta predominância de filotipos restritos a apenas um dos pacientes (77 %). Esta variabilidade na estrutura das comunidades bacterianas intra-radiculares também foi observada pelo emprego da técnica de T- RFLP em um grupo amostral mais heterogêneo (amostras assintomáticas, sensíveis e sintomáticas), ao se constatar que 50% dos fragmentos terminais de restrição (T-RFs) apresentaram-se exclusivos a, no máximo, 2 pacientes.

A técnica de T-RFLP foi aplicada na determinação da riqueza e abundância de componentes bacterianos da microbiota intra-radicular, revelando um total de 122 (enzima HhaI) e 123 (enzima MspI) fragmentos terminais de restrição (T-RFs) distintos. Se considerarmos que cada fragmento corresponde a, no mínimo, uma unidade taxonômica distinta (Marsh, 2005), pode-se afirmar que esta técnica permitiu a detecção de um mínimo de 123 filotipos bacterianos distintos. Este valor é significativamente maior que aquele obtido pela análise clonal (46 filotipos). Tal diferença foi também constatada quando se compararam as médias de unidades taxonômicas detectadas pelos dois estudos: 9,6 pela análise clonal, contra 20,4 por T-RFLP. Os diferentes tamanhos dos universos amostrais (7 versus 25) podem ter contribuído para a constatação destas diferenças. No entanto, como os primers 16S rDNA utilizados nas duas técnicas não foram os mesmos, parte das diferenças observadas pode ser explicada também por possíveis diferenças nas abrangências de detecção destes marcadores.

Uma parcela dos gêneros bacterianos identificados (preditos) pela análise em T-RFLP foi corroborada pela análise clonal como, por exemplo, Bacteroides, Burkholderia, Campylobacter, Dialister, Eubacterium, Peptostreptococcus, Selenomonas e Veillonella. Algumas exceções foram

51 constatadas, como Actinomyces, Bergeyella, Capnocytophaga, Mogibacterium, Peptococcus, identificados somente por T-RFLP e Catabacter, Desulfobulbus, Filifactor, Lactobacillus, Megasphaera e Treponema, detectados somente pela análise clonal. Visto que as duas técnicas possuem sensibilidades e resoluções taxonômicas distintas (Danovaro et al., 2007) e devido ao fato de diferentes marcadores universais haverem sido empregados, os resultados obtidos pelas duas técnicas devem ser, preferencialmente, considerados em conjunto, e não contrastados.

O emprego de primers espécie-específicos direcionados aos fatores de patogenicidade rgp e bspA em PCR em Tempo Real permitiu a detecção de P. gingivalis e T. forsythia em 24% e 56% dos pacientes, respectivamente. Entretanto, nenhuma destas espécies foi observada pela análise clonal, enquanto que apenas T. forsythia pôde ser predita por T-RFLP. Diferenças significativas são, em realidade, esperadas quando se empregam diferentes técnicas de detecção em amostras intra-radiculares (Gomes et al., 2005; Siqueira et al., 2002; Rolph et al., 2001). Realmente, a técnica de PCR em Tempo Real possui sensibilidade de detecção muito superior àquelas obtidas pelo cultivo ou pela PCR convencional, sendo capaz de diferenciar uma amostra negativa de outra contendo apenas 1 cópia do DNA-alvo (Morrison et al., 1988). É possível, contudo, que o uso de primers espécie-específicos tenha contribuído substancialmente na detecção de P. gingivalis e T. forsythia, visto que os primers 16S rDNA universais, comumente utilizados em estudos de detecção de bactérias orais, não englobam todos os grupos bacterianos (Horz et al., 2005).

Levando-se em consideração os resultados das três metodologias aplicadas nesta tese, pode-se afirmar que as técnicas moleculares independentes de cultivo demonstraram alta aplicabilidade no estudo das comunidades bacterianas associadas às infecções endodônticas. Em termos gerais, a análise de bibliotecas clonais de 16S rDNA possibilitou a detecção de novos filotipos da ordem Clostridiales, a PCR em Tempo Real demonstrou alta sensibilidade na detecção espécie-específica de dois importantes patógenos orais, enquanto que T-RFLP permitiu a caracterização global das estruturas bacterianas de forma abrangente e eficiente. Apesar de diferentes conjuntos de primers 16S rDNA haverem sido utilizados, notou-se uma relativa concordância entre as técnicas de análise de bibliotecas clonais e de T-RFLP, na medida em que demonstraram a existência de uma alta variabilidade na composição das comunidades bacterianas intra-radiculares e a detecção de unidades taxonômicas reconhecidamente endodônticas.

52 CONCLUSÕES

Com base nos resultados obtidos, pode-se concluir que:

1) A microbiota associada às infecções endodônticas é essencialmente polimicrobiana e composta, predominantemente, por bactérias anaeróbias Gram-positivas do filo Firmicutes.

2) Existe uma alta variabilidade na composição das comunidades bacterianas intra-radiculares entre diferentes indivíduos, tanto em relação à riqueza de espécies, como em relação aos tipos taxonômicos presentes.

3) Muito embora P. gingivalis e T. forsythia encontrem-se em consideráveis prevalências nas amostras intra-radiculares, os seus respectivos níveis celulares não apresentam correlação significativa com o desenvolvimento de sintomatologia dolorosa. O mesmo foi observado com respeito aos níveis celulares integrados (coexistência) de ambas as espécies.

4) A ausência de agrupamentos definidos de perfis bacterianos, segundo os parâmetros sintomatológicos, sugere que a estrutura das comunidades bacterianas intra-radiculares, em termos globais, não possui influência significativa no desenvolvimento da dor ou da sensibilidade de origem endodôntica.

Em suma, os resultados evidenciam que as técnicas moleculares permitem uma visão globalizada da composição microbiana que dificilmente seria obtida por técnicas tradicionais e que, portanto, possuem alta aplicabilidade na caracterização das comunidades bacterianas associadas às infecções endodônticas. A alta diversidade na estrutura das comunidades observadas e a detecção de filotipos nunca relatados em estudos anteriores sugerem que a microbiota endodôntica ainda requer caracterização adicional por técnicas contemporâneas.

53 REFERÊNCIAS

Aas JA, Paster BJ, Stokes LN, Olsen I, Dewhirst FE. Defining the normal bacterial flora of the oral cavity. J Clin Microbiol. 2003; 43: 5721–32. Blackwood CB, Hudleston D, Zak DR, Buyer JS. Interpreting ecological diversity indices applied to terminal restriction fragment length polymorphism data: insights from simulated microbial communities. Appl Environ Microbiol. 2007; 73: 5276-83. Brown Jr LR, Rudolph Jr CE. Isolation and identification of microorganisms from unexposed canals of pulp-involved teeth. Oral Surg. 1957; 10: 1094-9. Bustin SA. Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. J Mol Endocrinol. 2000; 25: 169–93. Bystrom A, Sundqvist G. Bacteriologic evaluation of the efficacy of mechanical root canal instrumentation in endodontic therapy. Scand J Dent Res. 1981; 89: 321-8. Danovaro R, Luna GM, Dell’Ano A. Measuring species richness based on microbial community fingerprints: the emperor has no clothes. Appl Environ Microbiol. 2007; 73: 2399–01. Fabricius L, Dahlén G, Öhman AE, Möller JR. Predominant indigenous oral bacteria isolated from infected root canals after varied times of closure. Scand J Dent Res. 1982; 90: 134-44. Foschi F, Cavrini F, Montebugnoli L, Stashenko P, Sambri V, Prati C. Detection of bacteria in endodontic samples by polymerase chain reaction assays and association with defined clinical signs in Italian patients. Oral Microbiol Immunol. 2005; 20: 289-95. Fouad AF, Barry J, Caimano M, Clawson M, Zhu Q, Carver R, Hazlett K, Radolf JD. PCR-based identification of bacteria associated with endodontic infections. J Clin Microbiol. 2002; 40: 3223- 31. Gomes BPFA, Jacinto RC, Pinheiro ET, Sousa ELR, Zaia AA, Ferraz CCR, Souza-Filho FJ. Porphyromonas gingivalis, Porphyromonas endodontalis, Prevotella intermedia and Prevotella nigrescens in endodontic lesions detected by culture and by PCR. Oral Microbiol Immunol. 2005; 20: 211-5. Gomes BPFA, Pinheiro ET, Gadê-Neto CR, Souza ELR, Ferraz CCR, Zaia AA, Teixeira FB, Souza-Filho FJ. Microbiological examination of infected dental root canals. Oral Microbiol Immunol . 2004; 19: 71-6. Grossman LI. Bacteriologic status of periapical tissue in 150 cases of pulpless teeth. J Dent Res. 1959; 38: 101-4. Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitative PCR. Genome Res. 1996; 6: 986-94. Holt SC, Kesavalu L, Walker S, Genco CA. Virulence factors of Porphyromonas gingivalis. Periodontol 2000. 1999; 20: 168-238. 54 Hommez GMG, Verhelst R, Claeys G, Vaneechoutte M, De Moor RJG. Investigation of the effect of the coronal restoration quality on the composition of the root canal microflora in teeth with apical periodontitis by means of T-RFLP analysis. Int Endod J. 2004; 37: 819–27. Horz HP, Vianna ME, Gomes BPFA, Conrads G. Evaluation of universal probes and primer sets for assessing total bacterial load in clinical samples: general implications and practical use in endodontic antimicrobial therapy. J Clin Microbiol. 2005; 43: 5332-7. Inagaki S, Onishi S, Kuramitsu HK, Sharma A. Porphyromonas gingivalis vesicles enhance attachment, and the leucine-rich repeat BspA protein is required for invasion of epithelial cells by Tannerella forsythia. Infect Immun. 2006; 74: 5023–8. Kantz WE, Henry CA. Isolation and classification of anaerobic bacteria from intact pulp chambers of non-vital teeth in man. Arch Oral Biol. 1974; 19: 91-6. Kroes I, Lepp PW, Relman DA. Bacterial diversity within the human subgingival crevice. Proc Natl Acad Sci U S A. 1999; 96: 14547–52. Lin LM, Skribner JE, Gaengler P. Factors associated with endodontic treatment failures. J Endod. 1992; 18: 625-7. Liu WT, Marsh TL, Cheng H, Forney LJ. Characterization of microbial diversity by determining terminal restriction fragment length polymorphisms of genes encoding 16s rRNA. Appl Eviron Microbiol. 1997; 63: 4516-22. Marsh TL. Culture-independent microbial community analysis with terminal restriction fragment length polymorphism. Methods Enzymol. 2005; 397: 308-29. Martin FE, Nadkarni MA, Jacques NA, Hunter N. Quantitative microbiological study of human carious dentine by culture and real-time PCR: association of anaerobes with histopathological changes in chronic pulpitis. J Clin Microbiol. 2002; 40: 1698-704. Melville TH, Birch RH. Root canal and periapical floras of infected teeth. Oral Surg. 1967; 23: 93- 8. Morrison TB, Weis JJ, Wittwer CT. Quantification of low-copy transcripts by continuous SYBR Green I monitoring during amplification. Biotechniques 1998; 24: 954-8. Moyer C, Tiedje JM, Dobbs FC, Karl DM. A computer-simulated restriction fragment length polymorphism analysis of bacterial small-subunit rRNA genes: efficacy of selected tetrameric restriction enzymes for studies of microbial diversity in nature. Applied Environ. Microbiol. 1996; 62: 2501–7. Munson MA, Pitt-Ford T, Chong B, Weightman A, Wade WG. 2002. Molecular and cultural analysis of the microflora associated with endodontic infections. J Dent Res. 81; 761–6. Osborn AM, Moore ER, Timmis KN. An evaluation of terminal-restriction fragment length polymorphism (T-RFLP) analysis for the study of microbial community structure and dynamics. Environ Microbiol. 2000; 2: 39-50. 55 Paster BJ, Falkler WA, Enwonwu Jr CO, Idigbe EO, KO Savage, VA Levanos, Tamer MA, Ericson RL, Lau CN, Dewhirst FE. Prevalent bacterial species and novel phylotypes in advanced noma lesions. J Clin Microbiol. 2002; 40: 2187–91. Paster BJ, Boches SK, Galvin JL, Ericson RE, Lau CN, Levanos VA, Sahasrabudhe A, Dewhirst FE. Bacterial diversity in human subgingival plaque. J Bacteriol. 2001; 183: 3770–83. Rôças IN, Siqueira Jr JF, Santos KRN, Coelho AMA. “Red complex” (Bacteroides forsythus, Porphyromonas gingivalis, and Treponema denticola) in endodontic infections: a molecular approach. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2001; 91: 468-71. Rolph HJ, Lennon A, Riggio MP, Saunders WP, MacKenzie D, Coldero L, Bagg J. Molecular identification of microorganisms from endodontic infections. J Clin Microbiol. 2001; 39: 3282–9. Sakamoto M, Rôças IN, Siqueira Jr JF, Benno Y. Molecular analysis of bacteria in asymptomatic and symptomatic endodontic infections. Oral Microbiol Immunol. 2006; 21: 112–22. Sakamoto M, Huang Y, Ohnishi M, Umeda M, Ishikawa I, Benno Y. Changes in oral microbial profiles after periodontal treatment as determined by molecular analysis of 16S rRNA genes. J Med Microbiol. 2004; 53: 563–71. Sakamoto M, Takeuchi Y, Umeda M, Ishikawa I, Benno Y.Application of terminal RFLP analysis to characterize oral bacterial flora in saliva of healthy subjects and patients with periodontitis. J Med Microbiol. 2003; 52: 79–89. Siqueira Jr JF, Rôças IN, Uzeda M, Colombo AP, Santo KRN. Comparison of 16S rDNA-based PCR and checkerboard DNA–DNA hybridisation for detection of selected endodontic pathogens. J Med Microbiol. 2002; 51: 1090–6. Socransky SS, Haffajee AD, Cugini MA, Smith C, Kent Jr RL. Microbial complexes in subgingival plaque. J Clin Periodontol. 1998; 25: 134-44. Socransky SS, Smith C, Haffajee AD. Subgingival microbial profiles in refractory periodontal disease. J Clin Periodontol. 2002; 29: 260–268. Sundqvist G. Bacteriological studies of necrotic dental pulps [dissertation]. Sweden: University of Umea°; 1976. Sundqvist G. Ecology of the root canal flora. J Endod. 1992; 18: 427–30. Sundqvist G. Taxonomy, ecology, and pathogenicity of the root canal flora. Oral Surg Oral Med Oral Pathol. 1994; 78: 522–30. Takemoto T, Kurihara H, Dahlen G. Characterization of Bacteroides forsythus isolates. J Clin Microbiol. 1997; 35: 1378-81. Verner C, Lemaitre P, Daniel A, Giumelli B, Lakhssassi N, Sixou M. Carpegen real-time polymerase chain reaction vs. anaerobic culture for periodontal pathogen identification. Oral Microbiol Immunol. 2006; 21: 341-6.

56 Wade WG, Spratt DA, Dymock D, Weightman AJ. Molecular detection of novel anaerobic species in dentoalveolar abscesses. Clin Infect Dis. 1997; 25: 235–6. Yoneda M, Hirofuji T, Anan H, Matsumoto A, Hamachi T, Nakayama K, Maeda K. Mixed infection of Porphyromonas gingivalis and Bacteroides forsythus in a murine abscess model: involvement of gingipains in a synergistic effect. J Periodontal Res. 2001; 36: 237-43.

57 ANEXO 1

58

59 ANEXO 2

60 ANEXO 3

61 ANEXO 4

62 ANEXO 5

63